[{"data":1,"prerenderedAt":9356},["ShallowReactive",2],{"/framework/blog/ai-financial-report-analysis-app":3,"navigation-framework":937,"/framework/blog/ai-financial-report-analysis-blog":1976,"related-candidates":2601,"surr-/framework/blog/ai-financial-report-analysis":9353},{"id":4,"title":5,"author":6,"body":11,"description":922,"extension":923,"meta":924,"navigation":931,"path":932,"seo":933,"sitemap":934,"stem":935,"__hash__":936},"content/framework/blog/ai-financial-report-analysis.md","Financial Report Analysis with LiveAI™",{"name":7,"description":8,"img":9,"provider":10},"Pathway Community","Multiple authors","/assets/blog/avatars/pathway-community-av.png","s3",{"type":12,"value":13,"toc":892},"minimark",[14,19,27,30,33,40,63,68,96,100,114,117,121,125,128,166,170,174,178,184,188,191,194,197,207,211,214,217,221,224,228,260,264,267,300,304,307,311,315,318,321,325,328,332,338,344,350,355,360,364,367,372,378,381,386,396,399,402,407,412,415,419,422,448,451,455,458,461,483,488,491,580,588,619,625,627,630,634,637,642,645,650,653,658,661,666,669,702,706,709,712,742,746,830,834],[15,16,18],"h2",{"id":17},"introduction","Introduction",[20,21,22,23],"p",{},"Most finance work starts with a pile of documents: annual reports, quarterly results, investor decks, filings. They’re long, they change often, and the facts you need are scattered across pages and versions. The result? Hours spent scrolling, searching, and cross-checking numbers just to answer a simple question like ",[24,25,26],"em",{},"“How did margins change compared to last quarter?”",[20,28,29],{},"This blog introduces a different way forward: Financial report analysis with LiveAI™. Instead of digging through static PDFs, you ask plain-English questions and get clear answers, always grounded in the latest version of your reports.",[20,31,32],{},"Financial teams drown in PDFs: annual reports, investor presentations, SEC/SEBI filings, broker notes; each packed with metrics that move decisions. This post shows how to do financial report analysis with LiveAI™, so you can turn those documents into reliable, queryable insights in minutes not days.",[20,34,35,39],{},[36,37,38],"strong",{},"Why LiveAI™ matters:"," traditional RAG pipelines go stale the moment your corpus changes. Pathway treats static and streaming data the same way: once your documents are preprocessed and indexed, Pathway auto-detects updates in your document directory and refreshes the store, keeping every answer grounded in the latest facts. In other words, your RAG stays live by design: no manual re-indexing, no nightly jobs, no lag.",[20,41,42,43,46,47,50,51,54,55,58,59,62],{},"Under the hood, we pair this real-time substrate with an agentic workflow designed for finance. A ",[36,44,45],{},"decider"," routes simple vs. complex asks; ",[36,48,49],{},"query augmentation"," expands finance shorthand (EPS, EBITDA, ROI, CAGR) to boost recall; ",[36,52,53],{},"multi-agent retrieval"," pulls from internal knowledge, the web, and downloaded finance PDFs; a ",[36,56,57],{},"consolidation step"," resolves conflicts across sources; and ",[36,60,61],{},"guardrails"," (PII/profanity checks) keep outputs safe and compliant. The result is a pipeline built to handle noisy PDFs and contradictory narratives without melting down.",[64,65,67],"h3",{"id":66},"why-you-should-care","Why you should care:",[69,70,71,78,84,90],"ul",{},[72,73,74,77],"li",{},[36,75,76],{},"Speed:"," jump from 200-page PDFs to answers (revenue, margins, guidance shifts) in a single query.",[72,79,80,83],{},[36,81,82],{},"Accuracy:"," conflict-aware consolidation reduces hallucinations and reconciles PRs vs. filings vs. transcripts.",[72,85,86,89],{},[36,87,88],{},"Coverage:"," works across annual reports, quarterly results, regulatory filings, and investor decks.",[72,91,92,95],{},[36,93,94],{},"Compliance:"," guardrails help you avoid leaking sensitive identifiers while staying audit-friendly.",[64,97,99],{"id":98},"what-youll-learn-in-this-post","What you’ll learn in this post",[69,101,102,105,108,111],{},[72,103,104],{},"How LiveAI™ keeps your RAG real-time without extra ETL.",[72,106,107],{},"A finance-tuned, multi-agent architecture that understands jargon and multi-hop questions.",[72,109,110],{},"A PDF-first ingestion pattern for balance sheets, income statements, cash-flows, and risk sections.",[72,112,113],{},"Practical prompts, schemas, and deployment tips you can reuse for your own AI financial repo",[20,115,116],{},"Read on to see how to plug this into your workflow and ship insights your stakeholders can trust.",[15,118,120],{"id":119},"financial-report-analysis-with-liveai-a-multi-agent-workflow-with-response-resolution","Financial Report Analysis with LiveAI™: A Multi-Agent Workflow with Response Resolution",[64,122,124],{"id":123},"overview","Overview",[20,126,127],{},"The solution integrates advanced techniques to build a robust Retrieval-Augmented Generation (RAG) system:",[69,129,130,136,142,148,154,160],{},[72,131,132,135],{},[36,133,134],{},"Decider agent",": A decider classifies queries as either simple or complex. Simple queries, which rely on general knowledge or straightforward logic, are answered directly using the agent's internal knowledge, while complex queries requiring tools, data processing, or sub-questions are forwarded for further processing (see Figure 1).",[72,137,138,141],{},[36,139,140],{},"Query Augmentation",": Detects jargon using a dynamic repository and enriches queries via a Large Language Model(LLM) to enhance retrieval accuracy.",[72,143,144,147],{},[36,145,146],{},"Pathway Integration",": Utilizes Pathway APIs with Google Drive for scalable storage and automated indexing, supporting diverse data formats (e.g., PDFs, JSON).",[72,149,150,153],{},[36,151,152],{},"Multi-Agent Retrieval",": Implements tools for direct answers, internal knowledge retrieval, web search, and domain-specific retrieval (e.g., financial data in PDFs).",[72,155,156,159],{},[36,157,158],{},"Conflict Resolution",": Addresses conflicting information through iterative knowledge consolidation, conflict flagging, and generating reliable answers. Guardrails ensure ethical and secure outputs.",[72,161,162,165],{},[36,163,164],{},"User Interface",": Combines Flask and React for a responsive and low-latency interface. This system ensures accurate, scalable, and context-aware retrieval and response generation.",[64,167,169],{"id":168},"system-demonstration-ai-financial-report-analysis","System Demonstration: AI Financial Report Analysis",[171,172],"video-player",{"src":173},"https://www.youtube.com/watch?v=C2N_QuN_r9A",[64,175,177],{"id":176},"code-repository-complete-setup-usage-guide","Code Repository - Complete Setup & Usage Guide",[179,180],"link-card",{"href":181,"title":182,"website":183},"https://github.com/Kesav73/AI-Financial-Report-Analysis.git","AI Financial Report Analysis","GitHub",[64,185,187],{"id":186},"query-decomposition-and-augmentation","Query Decomposition and Augmentation",[20,189,190],{},"As new information is continually generated, user queries may include terms that carry different meanings depending on the context. Specifically, queries containing jargon accompanied by some context can be reinterpreted to provide more precise or alternative meanings using an LLM.",[20,192,193],{},"To address this use case, we focus on identifying jargon within the user's query. This is achieved using a semi-dynamic predefined store that serves as a repository for querying and managing jargon. New jargon encountered can be added to this store, ensuring its adaptability over time.",[20,195,196],{},"Once jargon is identified, the entire query is passed to the LLM, which generates an augmented version enriched with additional contextual information. This process ensures that the retriever aligns more closely with the user's intent, enhancing the relevance and accuracy of the retrieved information.",[198,199],"article-img",{":zoomable":200,"alt":201,"className":202,"provider":10,"src":206},"true","",[203,204,205],"w-full","mx-auto","max-w-2xl","/assets/blog/earnings-call-transcript-analysis-framework/figure-1.jpg",[64,208,210],{"id":209},"integration-with-pathway","Integration with Pathway",[20,212,213],{},"The integration of Pathway offers a flexible and versatile approach, leveraging its diverse range of APIs. For data storage, we chose Google Drive, primarily due to its adaptability and compatibility with our design requirements. This choice also aligns with our scalable approach, enabling seamless support for various data formats, such as PDF, JSON, and SQL, with minimal effort.",[20,215,216],{},"Moreover, Pathway's existing vector store architecture facilitates efficient data management. It's reactivity allows for easy and rapid additions, deletions, and updates, all of which are automatically incorporated into the system. Additionally, Pathway's automated indexing capabilities further enhance the efficiency of our workflow, ensuring smooth and reliable data retrieval.",[64,218,220],{"id":219},"multi-agent-retrieval-for-financial-report-analysis","Multi-Agent Retrieval for Financial Report Analysis",[20,222,223],{},"Initially, we adopted a divide-and-regain strategy for our approach. However, we discovered that this method resulted in incomplete integration of conflicting knowledge. To address this limitation, we explored an alternative approach, as outlined below.",[198,225],{":zoomable":200,"alt":201,"className":226,"provider":10,"src":227},[203,204,205],"/assets/blog/earnings-call-transcript-analysis-framework/figure-2.jpg",[69,229,230,236,242,248,254],{},[72,231,232,235],{},[36,233,234],{},"Internal Knowledge agent",": This works when the LLM possesses the information in its internal knowledge to answer the query which has been asked.",[72,237,238,241],{},[36,239,240],{},"Retrieval Agent",": We have set up the \"retrieve_documents\" tool to search for relevant documents from the internal knowledge base.",[72,243,244,247],{},[36,245,246],{},"Web Search Agent",": Here, we used the \"tavily_tool\" to perform a web search which effectively finds critical and relevant information related to our context, if real time financial information like stock prices etc. is required then \"google_search_tool\" shall also be used.",[72,249,250,253],{},[36,251,252],{},"Downloaded Documents agent",": This implementation introduces a novel approach focused on extracting domain-specific knowledge, specifically targeting the finance sector. It enables the retrieval and processing of PDF documents, as financial information from major companies is often stored in this format. Future scalability plans include integrating a web-list, allowing the system to expand and gather domain-specific knowledge from a variety of fields beyond finance depending upon the use case.",[72,255,256,259],{},[36,257,258],{},"Consolidation agent",": The Consolidation Agent collects information retrieved by the above agents, it then processes this data using an iterative consolidation tool to reconcile discrepancies and ensure consistency. The final output consists of the most reliable and coherent information, providing a comprehensive and accurate response.",[64,261,263],{"id":262},"resolving-conflicting-responses","Resolving Conflicting Responses",[20,265,266],{},"Once we get data from retrieval, we proceed to resolving issues arising from unreliable external sources through the integration of internal and external information through a series of steps.",[268,269,270,276,282,288,294],"ol",{},[72,271,272,275],{},[36,273,274],{},"Adaptive Internal Knowledge Generation",": The LLM generates passages based on its internal knowledge relevant to the given query, this acts as a data source going forward.",[72,277,278,281],{},[36,279,280],{},"Iterative Source-Aware Knowledge Consolidation",": Next, we consolidate and refine the knowledge pool by selectively iterating through the generated and retrieved passages obtained from various sources.",[72,283,284,287],{},[36,285,286],{},"Identifying Conflicts",": Passages which present conflicting information are separated and flagged as conflicting. This allows the model to evaluate the reliability of each conflicting source, preventing the combination of contradictory data. We have also found the answers with a confidence which can be even converted to a score later on in future work.",[72,289,290,293],{},[36,291,292],{},"Answer Finalization",": Finally, we generate the final, reliable answer from the consolidated knowledge pool. This is done through the consideration of multiple perspectives to provide a well-rounded response.",[72,295,296,299],{},[36,297,298],{},"Guardrails",": It has been noticed that sometimes the answer provided could contain information which is private/controversial. To prevent the generation of answers like this, we have implemented rail checks which ensured that such information is not included in the final response.",[64,301,303],{"id":302},"integration-with-user-interface","Integration with User Interface",[20,305,306],{},"For the development of our solution, we integrated Flask with React, ensuring a seamless and efficient design. Special attention was given to minimizing latency to provide a smooth user experience. The resulting user interface features a dynamic and intuitive design, making it easy to navigate and interact with our solution.",[15,308,310],{"id":309},"novelty-and-specific-use-cases","Novelty and Specific Use-Cases",[64,312,314],{"id":313},"use-cases","Use Cases",[20,316,317],{},"Our solution is specifically designed for the finance domain, addressing limitations in current solutions that typically rely on scraping raw text from web pages. These existing methods often miss out on structured and essential data found in documents like PDFs. Our approach leverages a multi-agent system that not only searches for and downloads PDF documents, such as financial report press releases, but also extracts meaningful insights from them, like tables, graphs, and key financial metrics.",[20,319,320],{},"For example, while many financial websites may display earnings reports or balance sheets in a web page's text, the most accurate and comprehensive information often resides in downloadable PDF documents. Our system efficiently retrieves these documents, extracts critical data such as revenue, profits, and year-over-year growth rates, and processes them to provide a more structured and insightful analysis. By focusing on these high-value sources, we ensure that the information we deliver is both up-to-date and relevant, providing a more accurate picture of a company's financial health.",[64,322,324],{"id":323},"jargon-expansion-in-question-augmentation","Jargon Expansion in Question Augmentation",[20,326,327],{},"Our approach introduces jargon expansion as a key enhancement to the Agentic RAG framework, ensuring that domain-specific abbreviations and technical terms are fully understood by the retrieval system. This improves the accuracy and relevance of retrieved information, especially in domains with complex jargon.\nOur method stands out by automatically expanding jargon within financial queries, enabling the system to better interpret terms like EPS (Earnings Per Share) or P/E ratio (Price-to-Earnings Ratio). This ensures precise document retrieval and improves query understanding without requiring manual intervention.",[64,329,331],{"id":330},"use-cases-in-finance","Use Cases in Finance",[20,333,334,337],{},[36,335,336],{},"Stock Market Analysis",": Financial queries that contain shorthand terms like EPS or P/E ratio are expanded to their full forms, ensuring accurate retrieval of relevant stock market data.\nExample: \"What's the impact of EPS on stock performance?\" becomes \"What's the impact of Earnings Per Share (EPS) on stock performance?\"",[20,339,340,343],{},[36,341,342],{},"Investment Strategy",": Terms like ROI (Return on Investment) or CAGR (Compound Annual Growth Rate) are expanded to help retrieve more specific and relevant investment insights.\nExample: \"What's the effect of ROI on business growth?\" becomes \"What's the effect of Return on Investment (ROI) on business growth?\"",[20,345,346,349],{},[36,347,348],{},"Financial Reports",": Queries involving financial statements or metrics are enhanced by expanding abbreviations like EBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization), ensuring proper understanding in reports.\nExample: \"How does EBITDA affect company valuation?\" becomes \"How does Earnings Before Interest, Taxes, Depreciation, and Amortization (EBITDA) affect company valuation?\"",[20,351,352],{},[36,353,354],{},"Advantage",[69,356,357],{},[72,358,359],{},"This approach allows for easy domain-specific tuning without retraining models. By expanding jargon, we can fine-tune the system for the finance domain, ensuring efficient and precise query handling for financial analysis, investment strategies, and reporting.",[64,361,363],{"id":362},"question-decomposition","Question decomposition",[20,365,366],{},"In finance, users often ask questions that require insights from multiple angles. Query decomposition breaks down these complex questions into manageable sub-queries, enabling the system to retrieve and combine information from various sources for a more comprehensive answer.",[20,368,369],{},[36,370,371],{},"Example of Query Decomposition",[373,374,375],"blockquote",{},[20,376,377],{},"Original Query: \"Who is the CEO of the company that made the biggest loss in Q3 2024?\" This query requires two distinct pieces of information: The company that made the biggest loss. The CEO of that company.",[20,379,380],{},"To break it down:",[373,382,383],{},[20,384,385],{},"Sub-query 1: \"Which company made the biggest loss in Q3 2024?\"",[373,387,388],{},[20,389,390,391,395],{},"Sub-query 2: \"Who is the CEO of ",[392,393,394],"span",{},"Company Name","?\"",[20,397,398],{},"By splitting the query, the system can pull the relevant details from separate documents---one about Q3 2024 financial results and another listing company executives---and then combine them to generate the final answer.",[20,400,401],{},"Process in Action:",[373,403,404],{},[20,405,406],{},"Sub-query 1 Retrieval: Search for data on financial losses in Q3 2024 and find that Company XYZ reported the biggest loss.",[373,408,409],{},[20,410,411],{},"Sub-query 2 Retrieval: Search for information on Company XYZ and find that the CEO is Jane Doe.",[20,413,414],{},"Final Answer: \"The CEO of the company that made the biggest loss in Q3 2024, Company XYZ, is Jane Doe.\"",[64,416,418],{"id":417},"financial-data-from-pdfs-a-novel-approach-to-domain-specific-retrieval","Financial Data from PDFs: A Novel Approach to Domain-Specific Retrieval",[20,420,421],{},"A standout feature of our multi-agent retrieval system is its ability to extract and process financial data from PDFs. In industries like finance, crucial information is often stored in PDF format, especially when companies release detailed financial reports, earnings statements, regulatory filings, and annual reports. Traditional information retrieval systems might struggle to access and interpret this rich, domain-specific content.\nOur solution addresses this challenge by integrating a dedicated PDF extraction tool tailored specifically for financial data. This tool not only locates and retrieves PDF documents containing valuable financial knowledge but also processes them effectively to extract key insights such as:",[69,423,424,430,436,442],{},[72,425,426,429],{},[36,427,428],{},"Balance Sheets",": Extracting financial data related to assets, liabilities, and equity to understand a company's financial health.",[72,431,432,435],{},[36,433,434],{},"Income Statements",": Identifying revenue, expenses, and profit margins to assess operational performance.",[72,437,438,441],{},[36,439,440],{},"Cash Flow Statements",": Capturing cash inflows and outflows, which is crucial for understanding liquidity and financial stability.",[72,443,444,447],{},[36,445,446],{},"Investment Insights",": Parsing through investment-related documents, including corporate announcements and projections, that influence market behavior.",[20,449,450],{},"he financial PDF extraction tool operates seamlessly alongside other retrieval agents, ensuring that when users ask finance-related questions, the system can efficiently retrieve and process the most relevant, up-to-date financial information directly from these sources.\nFor example, if a user asks, \"What were the revenue trends of XYZ Corp in the Q4 report for 2024?\", the system would locate the latest PDF document containing XYZ Corp's financial report, extract the revenue data from the income statement, and present the insights directly in response.",[64,452,454],{"id":453},"filtering-out-relevant-information","Filtering Out Relevant Information",[20,456,457],{},"A key challenge in Retrieval-Augmented Generation (RAG) systems is ensuring the reliability and relevance of the information pulled from both internal and external sources. To address this, our approach integrates a series of systematic steps designed to enhance the performance of RAG while filtering out unreliable or conflicting data. This novel methodology ensures that the generated responses are both accurate and comprehensive.",[20,459,460],{},"Our information filtering process includes the following steps:",[69,462,463,468,473,478],{},[72,464,465,467],{},[36,466,274],{},": Initially, the LLM generates relevant passages from its internal knowledge base. These passages are generated based on the specific context of the query, ensuring that the information provided aligns with the user's needs and the domain of the query.",[72,469,470,472],{},[36,471,280],{},": Once the internal knowledge is generated, the system consolidates and refines the knowledge pool by iterating through both generated and retrieved passages. This step ensures that the most relevant and high-quality information is retained while less relevant data is filtered out.",[72,474,475,477],{},[36,476,286],{},": In the case where different passages provide conflicting information (such as contradictory facts or varying perspectives), these conflicting passages are flagged and separated. This step is crucial in preventing the combination of contradictory data in the final answer. The system evaluates the reliability of each conflicting source, ensuring that only the most trustworthy and consistent information contributes to the final response.",[72,479,480,482],{},[36,481,292],{},": Finally, the system generates the most reliable answer by drawing from the consolidated knowledge pool. The final response is shaped by considering multiple perspectives and integrating the best sources of information, ensuring that the result is both balanced and well-rounded.",[20,484,485],{},[36,486,487],{},"Advantages",[20,489,490],{},"This filtering process significantly improves the quality and accuracy of the generated answers. By systematically filtering and consolidating information from various sources, we ensure that the final response reflects the most reliable and relevant data. Moreover, by flagging conflicting information and evaluating its reliability, our system reduces the risk of propagating errors and inconsistencies, which is especially important in domains like finance, where accurate, dependable information is paramount.\nThrough this innovative filtering mechanism, we not only enhance the performance of RAG but also ensure that the responses are trustworthy, relevant, and comprehensive, providing users with a robust tool for answering complex queries.",[492,493,498,499,498,545],"table",{"className":494},[495,203,496,497],"table-border","table-full","text-center","\n  ",[500,501,502,503,502,525,498],"thead",{},"\n    ",[504,505,506,507,506,512,506,516,506,519,506,522,502],"tr",{},"\n      ",[508,509,511],"th",{"rowSpan":510},2,"Dataset",[508,513,515],{"colSpan":514},"2","Norm Rouge-1",[508,517,518],{"colSpan":514},"Norm Rouge-2",[508,520,521],{"colSpan":514},"Embed Rouge-1",[508,523,524],{"colSpan":514},"Embed Rouge-2",[504,526,506,527,530,506,533,535,506,537,539,506,541,543,502],{},[508,528,529],{},"Vanilla",[508,531,532],{},"R2R",[508,534,529],{},[508,536,532],{},[508,538,529],{},[508,540,532],{},[508,542,529],{},[508,544,532],{},[546,547,502,548,498],"tbody",{},[504,549,506,550,506,556,559,506,562,565,506,568,571,506,574,577,502],{},[551,552,555],"td",{"className":553,"rowSpan":510},[554],"font-bold","FinanceBench",[551,557,558],{},"0.0769",[551,560,561],{},"0.1334",[551,563,564],{},"0.0231",[551,566,567],{},"0.0426",[551,569,570],{},"0.0908",[551,572,573],{},"0.1663",[551,575,576],{},"0.0458",[551,578,579],{},"0.0899",[20,581,582],{},[24,583,587],{"className":584},[585,203,497,586],"block","-mt-4","Table 1: Rouge",[492,589,498,591,498,606],{"className":590},[495,203,496,497],[500,592,502,593,502,600,498],{},[504,594,506,595,506,597,502],{},[508,596,511],{"rowSpan":510},[508,598,599],{"colSpan":514},"Meteor",[504,601,506,602,604,502],{},[508,603,529],{},[508,605,532],{},[546,607,502,608,498],{},[504,609,506,610,506,613,616,502],{},[551,611,555],{"className":612,"rowSpan":510},[554],[551,614,615],{},"0.1069",[551,617,618],{},"0.1548",[20,620,621],{},[24,622,624],{"className":623},[585,203,497,586],"Table 2: Meteor Scores",[64,626,298],{"id":61},[20,628,629],{},"To ensure user safety and maintain ethical standards, we have implemented robust guardrails within the multi-agent RAG system. These guardrails are designed to identify and redact sensitive information such as credit card numbers, Aadhaar numbers, PAN, CVV, GSTIN, and IFSC codes using a combination of Presidio Analyzer and Presidio Anonymizer libraries. Additionally, we utilize Better Profanity to detect and censor offensive language, replacing inappropriate content with redaction markers. Custom rules tailored for India-specific identifiers enhance the system's ability to manage local regulatory requirements. By integrating these components, the system ensures that all outputs are sanitized, free from harmful or private information, and safe for user consumption. This critical feature reinforces the reliability and trustworthiness of the RAG system across diverse use cases.",[15,631,633],{"id":632},"evaluation-of-the-approach-for-financial-report-analysis","Evaluation of the Approach for Financial Report Analysis",[20,635,636],{},"The evaluation of our Retrieval-Augmented Generation (RAG) system focuses on assessing the effectiveness of the retrieval component, which plays a crucial role in selecting and ranking relevant documents or data. To gauge how well the retrieval phase operates, we measure its performance using several standard metrics, which allow us to monitor the precision and overall quality of our pipeline. In particular, we use metrics such as ROUGE scores and METEOR, calculated for different values of k, to provide insights into the effectiveness of our retrieval mechanism.",[20,638,639],{},[36,640,641],{},"ROUGE Score",[20,643,644],{},"ROUGE (Recall-Oriented Understudy for Gisting Evaluation) is a widely-used metric for evaluating automatic summarization and machine translation systems. It combines precision, recall, and F1-score by considering k-grams (subsequences of k words). For this evaluation, we have chosen ROUGE-1 and ROUGE-2 with values of k set to 1 and 2, respectively. The choice of these values comes from the observation that larger values of k often result in difficulty finding exact matching n-grams in scenarios where the answer can be augmented by the language model (LLM). Thus, using k = 1 and k = 2 allows for a reasonable balance between capturing meaningful content while acknowledging the challenges posed by larger k-grams.",[20,646,647],{},[36,648,649],{},"METEOR Score",[20,651,652],{},"METEOR (Metric for Evaluation of Translation with Explicit ORdering) is another automatic evaluation metric, originally developed for machine translation. It evaluates translation quality by comparing unigrams between machine-produced and human-produced reference translations. Unigrams are matched based on their surface forms, stemmed forms, and meanings. METEOR then calculates a score by combining unigram precision, unigram recall, and a measure of fragmentation that captures how well-ordered the matched words are relative to the reference. This ordering is particularly valuable in evaluating sentence accuracy and fluency.",[20,654,655],{},[36,656,657],{},"Datasets",[20,659,660],{},"Given that our solution is primarily focused on financial tasks, we selected datasets that are relevant to the domain of finance while also choosing general datasets to maintain comparability with other solutions. The datasets used for this evaluation are HOTPOTQA, TriviaQA, and FinanceBench (a dataset specifically focused on financial terms). These datasets allow us to assess the system's performance in both general knowledge and domain-specific contexts. For the evaluation, we present results from our RAG implementation on the FinanceBench dataset, compared to the baseline Vanilla RAG.",[20,662,663],{},[36,664,665],{},"Evaluation Results",[20,667,668],{},"The evaluation metrics used include both the ROUGE and METEOR scores, as shown in the Tables 1 and 2. Our results indicate a significant improvement in performance with our approach compared to Vanilla RAG.",[69,670,671,678,684,690,696],{},[72,672,673,674,677],{},"ROUGE-1: ",[36,675,676],{},"73.4%"," improvement.",[72,679,680,681,677],{},"ROUGE-2: ",[36,682,683],{},"84.7%",[72,685,686,687,677],{},"Embed ROUGE-1: ",[36,688,689],{},"83.0%",[72,691,692,693,677],{},"Embed ROUGE-2: ",[36,694,695],{},"96.4%",[72,697,698,699,677],{},"METEOR: ",[36,700,701],{},"44.8%",[15,703,705],{"id":704},"conclusion","Conclusion",[20,707,708],{},"This solution built using Pathway’s LiveAI™ framework addresses significant limitations in traditional RAG systems by introducing a multi-agent design capable of resolving conflicting responses and refining query decomposition. Its integration with Pathway and the ability to handle diverse data formats ensure scalability and adaptability. Focused on the finance domain, it demonstrates practical utility through its capacity to retrieve and analyze structured data, such as financial PDFs, with precision. The approach’s novelty lies in its modular design, conflict resolution mechanisms, and ability to expand into other domains. Future scalability through unified embeddings and Swanson linking shows promise for broader applications, making this solution a notable advancement in intelligent query handling and domain-specific retrieval systems.",[20,710,711],{},"If you are interested in diving deeper into the topic, here are some good references to get started with Pathway:",[69,713,714,721,727,735],{},[72,715,716],{},[717,718,720],"a",{"href":719},"/developers/user-guide/introduction/welcome","Pathway Developer Documentation",[72,722,723],{},[717,724,726],{"href":725},"/developers/templates","Pathway's Ready-to-run App Templates",[72,728,729],{},[717,730,734],{"href":731,"rel":732},"https://github.com/pathwaycom/llm-app/tree/main/templates/question_answering_rag",[733],"nofollow","End-to-end Real-time RAG app with Pathway",[72,736,737],{},[717,738,741],{"href":739,"rel":740},"https://discord.gg/pathway",[733],"Discord Community",[15,743,745],{"id":744},"references","References",[747,748,750,756],"callout",{"type":749},"basic",[751,752,753],"template",{"v-slot:summary":201},[20,754,755],{},"View Full List",[751,757,758],{"v-slot:content":201},[268,759,760,767,774,781,788,795,802,809,816,823],{},[72,761,762],{},[717,763,766],{"href":764,"rel":765},"https://arxiv.org/abs/2406.00029",[733],"Corrective Retrieval Augmented Generation — Shi-Qi Yan, Jia-Chen Gu, Yun Zhu & Zhen-Hua Ling",[72,768,769],{},[717,770,773],{"href":771,"rel":772},"https://arxiv.org/abs/2410.07176",[733],"Astute RAG: Overcoming Imperfect Retrieval Augmentation and Knowledge Conflicts for Large Language Models — Fei Wang, Xingchen Wan, Ruoxi Sun, Jiefeng Chen & Sercan O. Arık",[72,775,776],{},[717,777,780],{"href":778,"rel":779},"https://arxiv.org/abs/2404.16130",[733],"From Local to Global: A Graph RAG Approach to Query-Focused Summarization — Darren Edge, Ha Trinh, Newman Cheng, Joshua Bradley, Alex Chao, Apurva Mody, Steven Truitt & Jonathan Larson",[72,782,783],{},[717,784,787],{"href":785,"rel":786},"https://arxiv.org/abs/2408.00798",[733],"Golden-Retriever: High-Fidelity Agentic Retrieval Augmented Generation for Industrial Knowledge Base — Zhiyu An, Xianzhong Ding, Yen-Chun Fu, Cheng-Chung Chu, Yan Li, Wan Du",[72,789,790],{},[717,791,794],{"href":792,"rel":793},"https://www.deeplearning.ai/short-courses/langchain-for-llm-application-development/",[733],"Documentation for integrating Lanchain with llm",[72,796,797],{},[717,798,801],{"href":799,"rel":800},"https://arxiv.org/pdf/2401.15391",[733],"MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries — Yixuan Tang and Yi Yang",[72,803,804],{},[717,805,808],{"href":806,"rel":807},"https://arxiv.org/pdf/1809.09600",[733],"HOTPOTQA: A Dataset for Diverse, Explainable Multi-hop Question Answering",[72,810,811],{},[717,812,815],{"href":813,"rel":814},"https://arxiv.org/abs/1705.03551",[733],"TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension",[72,817,818],{},[717,819,822],{"href":820,"rel":821},"https://arxiv.org/abs/2311.11944",[733],"FinanceBench: A New Benchmark for Financial Question Answering",[72,824,825],{},[717,826,829],{"href":827,"rel":828},"https://aclanthology.org/W05-0909.pdf",[733],"METEOR: An Automatic Metric for MT Evaluation with Improved Correlation with Human Judgments",[15,831,833],{"id":832},"authors","Authors",[69,835,836,843,850,857,864,871,878,885],{},[72,837,838],{},[717,839,842],{"href":840,"rel":841},"https://www.linkedin.com/in/kesav-patneedi-b804932a9",[733],"Kesav Patneedi",[72,844,845],{},[717,846,849],{"href":847,"rel":848},"https://www.linkedin.com/in/rohan-kumar-mishra-996b911bb",[733],"Rohan Kumar Mishra",[72,851,852],{},[717,853,856],{"href":854,"rel":855},"https://www.linkedin.com/in/me-nishant-verma",[733],"Nishant Verma",[72,858,859],{},[717,860,863],{"href":861,"rel":862},"https://www.linkedin.com/in/bhavik-shangari-416b0324a",[733],"Bhavik Shangari",[72,865,866],{},[717,867,870],{"href":868,"rel":869},"https://www.linkedin.com/in/vedansh-sharma-202809218",[733],"Vedansh Sharma",[72,872,873],{},[717,874,877],{"href":875,"rel":876},"https://www.linkedin.com/in/uday-bhardwaj-054013265",[733],"Uday Bhardwaj",[72,879,880],{},[717,881,884],{"href":882,"rel":883},"https://www.linkedin.com/in/ojusgoel",[733],"Ojus Goel",[72,886,887],{},[717,888,891],{"href":889,"rel":890},"https://www.linkedin.com/in/avinash-patel-633195283",[733],"Avinash Patel",{"title":201,"searchDepth":510,"depth":510,"links":893},[894,899,909,918,919,920,921],{"id":17,"depth":510,"text":18,"children":895},[896,898],{"id":66,"depth":897,"text":67},3,{"id":98,"depth":897,"text":99},{"id":119,"depth":510,"text":120,"children":900},[901,902,903,904,905,906,907,908],{"id":123,"depth":897,"text":124},{"id":168,"depth":897,"text":169},{"id":176,"depth":897,"text":177},{"id":186,"depth":897,"text":187},{"id":209,"depth":897,"text":210},{"id":219,"depth":897,"text":220},{"id":262,"depth":897,"text":263},{"id":302,"depth":897,"text":303},{"id":309,"depth":510,"text":310,"children":910},[911,912,913,914,915,916,917],{"id":313,"depth":897,"text":314},{"id":323,"depth":897,"text":324},{"id":330,"depth":897,"text":331},{"id":362,"depth":897,"text":363},{"id":417,"depth":897,"text":418},{"id":453,"depth":897,"text":454},{"id":61,"depth":897,"text":298},{"id":632,"depth":510,"text":633},{"id":704,"depth":510,"text":705},{"id":744,"depth":510,"text":745},{"id":832,"depth":510,"text":833},"AI financial report analysis with LiveAI™. Extract insights from PDFs, apply multi-agent RAG, jargon expansion, and guardrails for accurate finance decisions.","md",{"layout":925,"date":926,"thumbnail":927,"tags":929},"blog","2025-09-03",{"src":928,"provider":10},"/assets/blog/thumbnails/earnings-call-transcript-analysis-framework-th.png",[930],"community",true,"/framework/blog/ai-financial-report-analysis",{"title":5,"description":922},{"loc":932},"framework/blog/ai-financial-report-analysis","zcNCW6Ij1AWNqBkH_t5g1OaZxCmCEFYnU2kzWg-niWw",[938],{"title":939,"path":940,"stem":941,"children":942,"page":951},"Framework","/framework","framework",[943],{"title":944,"path":945,"stem":946,"children":947,"meta":1975},"Blog","/framework/blog","framework/blog/1.index",[948,952,962,971,997,1010,1030,1046,1058,1068,1087,1096,1105,1118,1128,1143,1158,1167,1176,1190,1199,1210,1220,1229,1240,1251,1262,1271,1280,1289,1298,1310,1319,1328,1337,1356,1369,1383,1395,1404,1413,1422,1431,1442,1454,1463,1472,1481,1500,1514,1523,1532,1541,1549,1558,1572,1582,1591,1600,1609,1618,1627,1636,1645,1654,1663,1672,1682,1691,1700,1708,1717,1726,1735,1744,1753,1762,1772,1781,1790,1794,1803,1812,1821,1832,1840,1849,1857,1866,1883,1892,1907,1916,1925,1940,1949,1958,1966],{"title":944,"path":945,"stem":946,"meta":949},{"layout":950,"aside":951,"toc":951,"single":931},"default",false,{"title":953,"path":954,"stem":955,"meta":956},"Pathway is now in Open Beta","/framework/blog/pathway-open-beta-announced","framework/blog/1.pathway-open-beta-announced",{"layout":925,"thumbnail":957,"tags":959,"date":961,"hidden":931},{"src":958,"provider":10},"/assets/blog/thumbnails/pathway-thumbnail.gif",[925,960],"open beta","2022-12-05",{"title":963,"path":964,"stem":965,"meta":966},"Pathway named 2021 i-Lab Laureate","/framework/blog/ilab2021","framework/blog/1000.Ilab2021",{"layout":925,"aside":951,"thumbnail":967,"tags":969,"date":970,"hidden":931},{"src":968,"provider":10,"contain":931},"/assets/blog/thumbnails/I-Lab-LOGO.jpg",[925],"2021-07-09",{"title":972,"path":973,"stem":974,"meta":975},"Multimodal RAG with Gemini","/framework/blog/gemini-rag","framework/blog/1001.gemini-rag",{"aside":931,"layout":925,"thumbnail":976,"date":979,"tags":980,"keywords":983,"notebook_export_path":995,"run_template":996,"hidden":931},{"src":977,"fit":978},"/assets/content/showcases/gemini_rag/Blog_Banner.png","contain","2024-08-06",[981,982],"showcase","llm",[984,985,986,987,988,989,990,991,992,993,994],"LLM","RAG","GPT","OpenAI","Gemini","multimodal RAG","MM-RAG","unstructured","notebook","Gemini RAG","RAG Gemini","notebooks/showcases/multimodal-rag-using-Gemini.ipynb","/developers/templates/rag/template-multimodal-rag",{"title":998,"path":999,"stem":1000,"meta":1001},"Langchain and Pathway: RAG Apps with always-up-to-date knowledge","/framework/blog/langchain-integration","framework/blog/1002.langchain-integration",{"layout":925,"date":1002,"hidden":931,"thumbnail":1003,"tags":1005,"notebook_export_path":1007,"keywords":1008},"2024-05-18",{"src":1004},"/assets/content/showcases/vectorstore/Langchain-Pathway.png",[981,982,1006],"engineering","notebooks/showcases/langchain-integration.ipynb",[984,985,986,987,1009,992],"LangChain",{"title":1011,"path":1012,"stem":1013,"children":1014,"meta":1026},"LlamaIndex and Pathway: RAG Apps with always-up-to-date knowledge","/framework/blog/llamaindex-pathway","framework/blog/1003.llamaindex-pathway",[1015],{"title":1011,"path":1012,"stem":1013,"meta":1016},{"layout":925,"date":1017,"thumbnail":1018,"tags":1020,"keywords":1022,"docker_github_link":1025,"hidden":931},"2024-01-12",{"src":1019},"/assets/content/showcases/vectorstore/llamaindexpathway.png",[981,982,1021,1006],"case-study",[984,985,986,987,1023,1024],"LlamaIndex","docker","https://github.com/pathway-labs/realtime-indexer-qa-chat/tree/main",{"layout":925,"date":1017,"thumbnail":1027,"tags":1028,"keywords":1029,"docker_github_link":1025,"hidden":931},{"src":1019},[981,982,1021,1006],[984,985,986,987,1023,1024],{"title":1031,"path":1032,"stem":1033,"meta":1034},"Real-time Enterprise RAG with SharePoint","/framework/blog/enterprise_rag_sharepoint","framework/blog/1004.enterprise_rag_sharepoint",{"aside":931,"layout":925,"thumbnail":1035,"tags":1037,"date":1038,"updatedDate":1039,"keywords":1040,"docker_github_link":731},{"src":1036,"fit":978},"/assets/content/showcases/enterprise_sharepoint_rag/Enterprise_RAG-thumbnail.png",[981,982,1021,1006],"2024-07-15","2025-03-24",[984,985,1041,1042,1043,1044,1045],"Real-time RAG","LiveAI™","Enterprise RAG","Docker","SharePoint",{"title":1047,"path":1048,"stem":1049,"meta":1050},"Evaluating RAG applications with RAGAS","/framework/blog/evaluating-rag","framework/blog/1005.evaluating-rag",{"layout":925,"date":1051,"thumbnail":1052,"tags":1054,"notebook_export_path":1055,"keywords":1056,"hidden":931},"2025-03-13",{"src":1053},"/assets/content/showcases/evals/rag-evals.png",[981,982,1006],"notebooks/tutorials/rag-evaluations.ipynb",[984,985,986,987,1057,992],"evaluations",{"title":1059,"path":1060,"stem":1061,"meta":1062},"How to use ChatGPT API in Python for your real-time data","/framework/blog/chatgpt-python-api-real-time-data","framework/blog/1010.chatgpt-python-api-real-time-data",{"aside":951,"layout":925,"thumbnail":1063,"tags":1065,"date":1067,"related":951,"hidden":931},{"src":1064,"provider":10},"/assets/blog/thumbnails/chatgpt-python-api-real-time-data.png",[925,1066,1006],"tutorial","2023-08-28",{"title":1069,"path":1070,"stem":1071,"meta":1072},"How La Poste uses Pathway microservices to deliver high-quality ETAs","/framework/blog/pathway-laposte-microservices","framework/blog/1020.pathway-laposte-microservices",{"layout":751,"thumbnail":1073,"date":1075,"tags":1076,"keywords":1078,"hidden":931},{"src":1074,"provider":10,"contain":931},"/assets/blog/thumbnails/pathway-laposte-microservices-th.png","2025-05-15",[981,1077],"data-pipeline",[1079,1080,1081,1082,1083,1084,1085,1086],"ETL","Microservices","Microservice architecture","Micropipelines","Delta tables","S3","La Poste","ETA",{"title":1088,"path":1089,"stem":1090,"meta":1091},"50× Faster Local Embeddings with Batch UDFs","/framework/blog/local-embeddings-batch-udfs","framework/blog/1025.local-embeddings-batch-udfs",{"layout":925,"thumbnail":1092,"date":1094,"tags":1095,"hidden":931},{"src":1093,"provider":10,"contain":931},"/assets/blog/thumbnails/batch-udfs.png","2025-07-17",[925,1006],{"title":1097,"path":1098,"stem":1099,"meta":1100},"Real-Time OCR with PaddleOCR and Pathway","/framework/blog/paddleocr","framework/blog/699.paddleocr",{"single":931,"aside":951,"layout":925,"date":1101,"thumbnail":1102,"tags":1104},"2026-02-02",{"src":1103},"/assets/content/blog/thumbnails/paddleocr-thumbnail.png",[925,1066,1006,941],{"title":1106,"path":1107,"stem":1108,"meta":1109},"Build LLM/RAG pipelines with YAML templates by Pathway","/framework/blog/llm-yaml-templates","framework/blog/700.llm-yaml-templates",{"layout":925,"date":1110,"tags":1111,"thumbnail":1113,"related":1115,"hidden":931},"2024-10-29",[1112,1006],"news",{"src":1114,"provider":10},"/assets/blog/thumbnails/llm-yaml-templates-th.png",[1116,1117],"/blog/pathway-showcased-during-intel-ai-summit","/blog/pathway-joins-the-opea-to-accelerate-generative-ai-adoption",{"title":1119,"path":1120,"stem":1121,"meta":1122},"Pathway is Now Available on Microsoft Azure!","/framework/blog/azure-aci-deploy","framework/blog/701.azure-aci-deploy",{"layout":925,"date":1123,"tags":1124,"thumbnail":1125,"related":1127,"hidden":931},"2024-11-27",[1112],{"src":1126,"provider":10},"/assets/blog/thumbnails/pathway-azure-marketplace-th.png",[1116,1117],{"title":1129,"path":1130,"stem":1131,"meta":1132},"Gemini 2.0 for Document Ingestion and Analytics with Pathway","/framework/blog/gemini2-document-ingestion-and-analytics","framework/blog/878.gemini2-document-ingestion-and-analytics",{"layout":925,"date":1133,"thumbnail":1134,"tags":1136,"coauthors":1137,"hidden":931},"2025-02-20",{"src":1135,"provider":10},"/assets/blog/thumbnails/gemini-2-th.png",[925,1006],[1138],{"name":1139,"description\"":1140,"img":1141,"provider":10,"linkedin":1142},"Berke Can Rizai","LLM Research Engineer","/assets/pictures/image_pathway_team.png","https://www.linkedin.com/in/berke-can-rizai/",{"title":1144,"path":1145,"stem":1146,"meta":1147},"Pathway’s Apache Iceberg Connectors for Real-Time Data Pipelines","/framework/blog/pathway-apache-iceberg-connectors","framework/blog/879.pathway-apache-iceberg-connectors",{"layout":925,"date":1148,"thumbnail":1149,"tags":1151,"coauthors":1152,"hidden":931},"2025-02-11",{"src":1150,"provider":10},"/assets/blog/thumbnails/azure-iceberg-th.png",[925,1066],[1153],{"name":1154,"description":1155,"img":1156,"provider":10,"linkedin":1157},"Sergey Kulik","Lead Software Research Engineer at Pathway","/assets/pictures/image_kulik_pathway.jpg","https://www.linkedin.com/in/sergey-kulik-72506a33/",{"title":1159,"path":1160,"stem":1161,"meta":1162},"Real-Time AI Pipeline with DeepSeek, Ollama and Pathway","/framework/blog/deepseek-ollama","framework/blog/880.deepseek-ollama",{"single":931,"aside":951,"layout":925,"date":1163,"thumbnail":1164,"tags":1166,"hidden":931},"2025-02-05",{"src":1165,"provider":10},"/assets/blog/thumbnails/real-time-rag-pipeline-th.png",[925,1066,1006],{"title":1168,"path":1169,"stem":1170,"meta":1171},"Power and Deploy RAG Agent Tools with Pathway","/framework/blog/deploy-rag-agent-tools-with-pathway","framework/blog/881.deploy-rag-agent-tools-with-pathway",{"layout":925,"date":1172,"thumbnail":1173,"tags":1175,"hidden":931},"2025-01-16",{"src":1174,"provider":10},"/assets/blog/thumbnails/power-and-deploy-th.png",[925,1006],{"title":1177,"path":1178,"stem":1179,"meta":1180},"How Text Embeddings help suggest similar words","/framework/blog/how-text-embeddings-help-suggest-similar-words","framework/blog/881.how-text-embeddings-help-suggest-similar-words",{"layout":925,"date":1148,"thumbnail":1181,"tags":1183,"coauthors":1184},{"src":1182,"provider":10},"/assets/blog/thumbnails/pathway-text-embeddings-th.png",[930],[1185],{"name":1186,"description":1187,"img":1188,"provider":10,"linkedin":1189},"Yashasvee Taiwade","Student at Indian Institute of Technology, Bombay","/assets/blog/avatars/yashasvee-avatar.png","https://www.linkedin.com/in/yashasvee-taiwade-84826a337/",{"title":1191,"path":1192,"stem":1193,"meta":1194},"2025 PATHWAY CEO PREDICTIONS","/framework/blog/pathway-predictions-2025","framework/blog/882.pathway-predictions-2025",{"layout":925,"date":1195,"thumbnail":1196,"tags":1198,"hidden":931},"2024-12-20",{"src":1197,"provider":10},"/assets/blog/thumbnails/pathway-2025-predictions-th.png",[925],{"title":1200,"path":1201,"stem":1202,"meta":1203},"Scalable Alternative to Apache Kafka and Flink for Advanced Streaming: Build Real-Time Systems with NATS and Pathway","/framework/blog/build-real-time-systems-nats-pathway-alternative-kafka-flink","framework/blog/883.build-real-time-systems-nats-pathway-alternative-kafka-flink",{"layout":925,"date":1204,"thumbnail":1205,"tags":1207,"coauthors":1208,"hidden":931},"2024-12-11",{"src":1206,"provider":10},"/assets/blog/thumbnails/pathway-nats-th.png",[1066],[1209],{"name":1154,"description":1155,"img":1156,"linkedin":1157},{"title":1211,"path":1212,"stem":1213,"meta":1214},"Achieve Sub-Second Latency with your S3 Storage without Kafka","/framework/blog/kafka-alternative","framework/blog/884.kafka-alternative",{"layout":925,"redirect":1215,"date":1216,"thumbnail":1217,"tags":1219,"hidden":931},"/developers/templates/etl/kafka-alternative","2024-08-27",{"src":1218},"/assets/content/showcases/kafka-alternatives/kafka-alternative-banner.png",[1066,1006,1021],{"title":1221,"path":1222,"stem":1223,"meta":1224},"Computing the Option Greeks using Pathway and Databento","/framework/blog/option-greeks","framework/blog/885.option-greeks",{"layout":925,"redirect":1225,"date":979,"thumbnail":1226,"tags":1228,"hidden":931},"/developers/templates/etl/option-greeks",{"src":1227},"/assets/content/showcases/option-greeks/option-greeks.svg",[1066,1006],{"title":1230,"path":1231,"stem":1232,"meta":1233},"Harnessing the Power of Now With Real-Time Analytics with Zuzanna Stamirowska & Hélène Stanway","/framework/blog/datacamp-real-time-analytics","framework/blog/886.datacamp-real-time-analytics",{"layout":925,"redirection":931,"date":1234,"thumbnail":1235,"tags":1237,"hidden":931},"2024-11-13",{"src":1236},"https://i3.ytimg.com/vi/6LcRxim5W1Y/maxresdefault.jpg",[1238,1239],"video","podcast",{"title":1241,"path":1242,"stem":1243,"meta":1244},"Adaptive RAG: cut your LLM costs without sacrificing accuracy","/framework/blog/adaptive-rag","framework/blog/887.adaptive-rag",{"layout":925,"redirect":1245,"date":1246,"thumbnail":1247,"tags":1249,"hidden":931},"/developers/templates/rag/adaptive-rag","2024-03-28",{"src":1248},"/assets/content/blog/adaptive-rag-plots/visual-abstract.png",[925,1250,1006,1066,1021],"feature",{"title":1252,"path":1253,"stem":1254,"meta":1255},"Pathway Slide Search is now available on Intel Tiber Cloud","/framework/blog/intel-summit","framework/blog/890.intel-summit",{"layout":925,"date":1256,"thumbnail":1257,"tags":1259,"related":1260,"hidden":931},"2024-09-23",{"src":1258,"provider":10},"/assets/blog/thumbnails/intel-accelerator-program-th.png",[1112,1021],[1116,1261],"/blog/intel-liftoff",{"title":1263,"path":1264,"stem":1265,"meta":1266},"Pathway Joins Linux Foundation’s OPEA alongside Intel and HuggingFace","/framework/blog/pathway-joins-the-opea-to-accelerate-generative-ai-adoption","framework/blog/891.pathway-joins-the-opea-to-accelerate-generative-ai-adoption",{"layout":925,"date":1267,"thumbnail":1268,"tags":1270,"hidden":931},"2024-08-29",{"src":1269,"provider":10},"/assets/blog/thumbnails/pathway-joins-opea-th.png",[1112],{"title":1272,"path":1273,"stem":1274,"meta":1275},"Pathway showcased during Intel AI Summit","/framework/blog/pathway-showcased-during-intel-ai-summit","framework/blog/892.pathway-showcased-during-intel-ai-summit",{"layout":925,"date":1276,"thumbnail":1277,"tags":1279,"hidden":931},"2024-09-02",{"src":1278,"provider":10},"/assets/blog/thumbnails/intel-ai-summit-blog-th.png",[1112],{"title":1281,"path":1282,"stem":1283,"meta":1284},"BNP Paribas Talk: AI and Real-time data processing in the banking industry, Vivatech 2024","/framework/blog/bnp-talk-at-vivatech","framework/blog/893.bnp-talk-at-vivatech",{"layout":925,"date":1285,"thumbnail":1286,"tags":1288,"hidden":931},"2024-05-22",{"src":1287,"provider":10},"/assets/blog/thumbnails/bnp-talk-at-vivatech-th.png",[1238,1239],{"title":1290,"path":1291,"stem":1292,"meta":1293},"Pathway at Modern Data Stack Meetup: Enabling Real-Time Operational Analytics for La Poste Groupe","/framework/blog/modern-data-network","framework/blog/894.modern-data-network",{"layout":925,"date":1294,"thumbnail":1295,"tags":1297,"hidden":931},"2024-01-31",{"src":1296,"provider":10},"/assets/blog/thumbnails/modern-data-network-th.png",[1112,1021],{"title":1299,"path":1300,"stem":1301,"meta":1302},"Transdev Project kickoff in Dunkirk, France","/framework/blog/project-kickoff-in-dunkirk","framework/blog/895.project-kickoff-in-dunkirk",{"layout":925,"date":1303,"thumbnail":1304,"tags":1306,"related":1307,"hidden":931},"2024-01-26",{"src":1305,"provider":10},"/assets/blog/thumbnails/project-kickoff-in-dunkirk-th.png",[1112,1021],[1308,1309],"/blog/intel-accelerator-program","/blog/innovation-challenge",{"title":1311,"path":1312,"stem":1313,"meta":1314},"Pathway has been selected for Intel Liftoff","/framework/blog/intel-liftoff","framework/blog/896.intel-liftoff",{"layout":925,"date":1315,"thumbnail":1316,"tags":1317,"related":1318,"hidden":931},"2024-06-01",{"src":1258,"provider":10},[1112],[1308,1309],{"title":1320,"path":1321,"stem":1322,"meta":1323},"Pathway has been selected for the Intel’s accelerator program for deep tech startups","/framework/blog/intel-accelerator-program","framework/blog/897.intel-accelerator-program",{"layout":925,"date":1324,"thumbnail":1325,"tags":1326,"related":1327,"hidden":931},"2024-04-17",{"src":1258,"provider":10},[1112],[1309],{"title":1329,"path":1330,"stem":1331,"meta":1332},"Finalist of the NATO Innovation Challenge!","/framework/blog/innovation-challenge","framework/blog/899.innovation-challenge",{"layout":925,"date":1333,"thumbnail":1334,"tags":1336,"hidden":931},"2023-11-14",{"src":1335,"provider":10},"/assets/blog/thumbnails/innovation-challenge-th.png",[1112],{"title":1338,"path":1339,"stem":1340,"meta":1341},"Realtime Classification with Nearest Neighbors","/framework/blog/realtime-classification","framework/blog/900.realtime-classification",{"layout":925,"date":1342,"thumbnail":1343,"tags":1345,"keywords":1347,"hidden":931},"2022-10-26",{"src":1344,"provider":10},"/assets/blog/thumbnails/pathway-card.png",[1066,1346],"machine-learning",[1348,1349,1350,1351,1352,1353,1354,1355],"Classification","regression","KNN","LSH","index","Locality Sensitive Hashing","MNIST","euclidean distance",{"title":1357,"path":1358,"stem":1359,"meta":1360},"Building End-to-End RAG with NPCI’s AI Leader","/framework/blog/building-end-to-end-rag-from-hardware-to-application","framework/blog/901.building-end-to-end-rag-from-hardware-to-application",{"keywords":1361,"layout":925,"thumbnail":1365,"tags":1367,"date":1368,"hidden":931},[1362,1363,985,1364],"data preprocessing","adaptive RAG","similarity search",{"src":1366,"provider":10},"/assets/blog/thumbnails/building-end-to-end-rag-from-hardware-to-app-th.png",[925,1021],"2024-04-29",{"title":1370,"path":1371,"stem":1372,"meta":1373},"Machine Unlearning for LLMs: Build Apps that Self-Correct in Real-Time ","/framework/blog/machine-unlearning","framework/blog/902.machine-unlearning",{"keywords":1374,"layout":925,"tags":1379,"date":1380,"thumbnail":1381,"hidden":931},[1375,1376,1377,1363,985,1378,1352],"machine unlearning","LLMs","chatbot","adaptive index",[925],"2024-04-19",{"src":1382},"/assets/content/blog/ML-unlearning-forgotten.svg",{"title":1384,"path":1385,"stem":1386,"meta":1387},"Batch processing vs stream processing","/framework/blog/batch-processing-vs-stream-processing","framework/blog/903.batch-processing-vs-stream-processing",{"keywords":1388,"layout":925,"thumbnail":1392,"tags":1393,"date":1394,"hidden":931},[1389,1390,1391],"batch processing","stream processing","data streaming",{"src":1344,"provider":10},[925],"2024-04-18",{"title":1396,"path":1397,"stem":1398,"meta":1399},"The Power of Real-Time Visualization for Logistics IoT","/framework/blog/power-of-real-time-visualization-for-logistics","framework/blog/903.power-of-real-time-visualization-for-logistics",{"redirection":931,"thumbnail":1400,"tags":1402,"date":1403,"hidden":931},{"src":1401,"provider":10},"/assets/blog/thumbnails/power-of-real-time-visualization-for-logistics-th.png",[1021],"2024-04-16",{"title":1405,"path":1406,"stem":1407,"meta":1408},"Supply Chain Optimization with Integrated IoT Data","/framework/blog/supply-chain-optimization-with-integrated-iot-data","framework/blog/904.supply-chain-optimization-with-integrated-iot-data",{"redirection":931,"thumbnail":1409,"tags":1411,"date":1412,"hidden":931},{"src":1410,"provider":10},"/assets/blog/thumbnails/supply-chain-optimization-with-integrated-iot-data-th.png",[925],"2024-03-21",{"title":1414,"path":1415,"stem":1416,"meta":1417},"Build a real-time RAG chatbot using Google Drive and Sharepoint","/framework/blog/build-a-realtime-rag-chatbot-streamlit","framework/blog/905.build-a-realtime-rag-chatbot-streamlit",{"aside":951,"single":931,"redirection":931,"thumbnail":1418,"tags":1420,"date":1421,"hidden":931},{"src":1419,"provider":10},"/assets/blog/thumbnails/title_streamlit-pathway-llm-google-doc-th.svg",[925,1021],"2024-03-07",{"title":1423,"path":1424,"stem":1425,"meta":1426},"Redefining AI’s Learning Curve: The Art of Unlearning","/framework/blog/tech-talks-podcast","framework/blog/906.tech-talks-podcast",{"redirection":931,"thumbnail":1427,"tags":1429,"date":1430,"hidden":931},{"src":1428},"https://techblogwriter.co.uk/wp-content/uploads/2023/12/Zuzanna-Stamirowska-CEO-and-Co-Founder-of-Pathway-1.jpg",[1239],"2023-12-25",{"title":1432,"path":1433,"stem":1434,"meta":1435},"RAG - Streaming Application Pathway Legal Assistant","/framework/blog/rag-streaming-app-pathway-legacy-assistant","framework/blog/907.rag-streaming-app-pathway-legacy-assistant",{"layout":925,"keywords":1436,"thumbnail":1438,"tags":1439,"date":1441,"hidden":931},[1437],"rag",{"src":1344,"provider":10},[1440],"pathway","2023-12-29",{"title":1443,"path":1444,"stem":1445,"meta":1446},"Retrieval Augmented Generation: Beginner’s Guide to RAG Apps","/framework/blog/retrieval-augmented-generation-beginners-guide-rag-apps","framework/blog/908.retrieval-augmented-generation-beginners-guide-rag-apps",{"layout":925,"keywords":1447,"thumbnail":1451,"tags":1453,"date":1441,"hidden":931},[1437,1448,1449,1450],"rag llm","vector index","retrieval augmented generation",{"src":1452,"provider":10},"/assets/blog/thumbnails/rag-beginners-guide-th.png",[925,1437],{"title":1455,"path":1456,"stem":1457,"meta":1458},"Guest speaker on the Digitalisation World podcast","/framework/blog/digitalisation-world-podcast","framework/blog/909.digitalisation-world-podcast",{"redirection":931,"thumbnail":1459,"tags":1461,"date":1462,"hidden":931},{"src":1460,"provider":10},"/assets/blog/thumbnails/digitalisation-world-th.png",[1239],"2023-11-22",{"title":1464,"path":1465,"stem":1466,"meta":1467},"Guest speaker on The Futurists podcast","/framework/blog/futurist-podcast","framework/blog/910.futurist-podcast",{"layout":925,"aside":951,"thumbnail":1468,"tags":1470,"date":1471,"hidden":931},{"src":1469,"provider":10},"/assets/blog/thumbnails/thefuturists-th.webp",[925,1239],"2023-11-10",{"title":1473,"path":1474,"stem":1475,"meta":1476},"How to build a real-time LLM app without vector databases","/framework/blog/build-a-real-time-llm-app-without-vector-databases-using-pathway","framework/blog/911.build-a-real-time-llm-app-without-vector-databases-using-pathway",{"aside":951,"single":931,"redirection":931,"thumbnail":1477,"tags":1479,"date":1480,"hidden":931},{"src":1478,"provider":10},"/assets/blog/thumbnails/relatime-app-th.png",[925],"2023-10-19",{"title":1482,"path":1483,"stem":1484,"meta":1485},"Signal Processing in Real-time: Bridging the Gap Between Ideal Sampling and Real-World Data Streams","/framework/blog/signal-processing","framework/blog/912.signal-processing",{"keywords":1486,"layout":925,"thumbnail":1493,"tags":1495,"date":1496,"related":1497,"hidden":931},[1487,1488,1489,1490,1491,1492],"time series","signal processing","resampling","filter","gaussian","upsampling",{"src":1494},"/assets/content/blog/iot-data-analytics/manul-iot.png",[925],"2023-10-16",[1498,1499],"/developers/templates/etl/gaussian_filtering_python","/developers/templates/etl/upsampling",{"title":1501,"path":1502,"stem":1503,"meta":1504},"Kafka vs RabbitMQ for Data Streaming","/framework/blog/kafka-vs-rabbit-mq-for-data-streaming","framework/blog/913.kafka-vs-rabbit-mq-for-data-streaming",{"keywords":1505,"layout":925,"thumbnail":1508,"tags":1510,"date":1511,"related":1512,"hidden":931},[1506,1507,1391],"kafka","rabbitmq",{"src":1509,"provider":10},"/assets/blog/thumbnails/rabbitmqvskafka-th.png",[925],"2023-09-29",[1513],"/blog/building-enterprise-search-apis-with-llms-for-production",{"title":1515,"path":1516,"stem":1517,"meta":1518},"Building LLM enterprise search APIs","/framework/blog/llm-enterprise-search","framework/blog/914.llm-enterprise-search",{"layout":925,"thumbnail":1519,"tags":1521,"date":1522,"hidden":931},{"src":1520,"provider":10},"/assets/blog/thumbnails/llm-with-search-apis-th.png",[925],"2023-08-23",{"title":1524,"path":1525,"stem":1526,"meta":1527},"Pathway: Fastest Data Processing Engine - 2023 Benchmarks","/framework/blog/streaming-benchmarks-pathway-fastest-engine-on-the-market","framework/blog/915.streaming-benchmarks-pathway-fastest-engine-on-the-market",{"layout":925,"thumbnail":1528,"tags":1530,"date":1531,"hidden":931},{"src":1529,"contain":931},"/assets/content/blog/benchmark/bm-preview.png",[925],"2023-07-17",{"title":1533,"path":1534,"stem":1535,"meta":1536},"Tutorial: Distributed computing with live streaming data","/framework/blog/distributed-computing-with-live-streaming-data","framework/blog/918.distributed-computing-with-live-streaming-data",{"layout":925,"thumbnail":1537,"tags":1539,"date":1540,"hidden":931},{"src":1538},"/assets/content/blog/distributed-computing-with-live-streaming-data-th.png",[925,1066],"2023-06-16",{"title":1542,"path":1543,"stem":1544,"meta":1545},"IoT Data Analytics: Processing Real-World Data in Real Time","/framework/blog/iot-data-analytics","framework/blog/918.iot-data-analytics",{"layout":925,"thumbnail":1546,"tags":1547,"date":1548,"hidden":931},{"src":1494},[925],"2023-06-21",{"title":1550,"path":1551,"stem":1552,"meta":1553},"Guest Speaker on Intel Business Podcast with Darren Pulsipher, Chief Solutions Architect","/framework/blog/intel-business-podcast","framework/blog/919.intel-business-podcast",{"layout":925,"thumbnail":1554,"tags":1556,"date":1557,"hidden":931},{"src":1555},"https://img.youtube.com/vi/Yn-EUssQG04/maxresdefault.jpg",[925,1239],"2023-05-23",{"title":1559,"path":1560,"stem":1561,"meta":1562},"Pathway is Featured in Gartner’s Market Guide for Event Stream Processing","/framework/blog/market-guide-event-stream-processing","framework/blog/920.market-guide-event-stream-processing",{"layout":925,"tags":1563,"date":1564,"thumbnail":1565,"enterprise":931,"keywords":1567,"hidden":931},[1112],"2023-05-15",{"src":1566,"contain":931,"padding":931},"/assets/content/blog/gartner-th.png",[1568,1440,1569,1570,1571],"gartner","event","stream","processing",{"title":1573,"path":1574,"stem":1575,"meta":1576},"Guest Speaker on Charbon Podcast (in French)","/framework/blog/charbon-podcast-with-claire","framework/blog/921.charbon-podcast-with-claire",{"layout":925,"thumbnail":1577,"tags":1579,"lang":1580,"date":1581,"hidden":931},{"src":1578},"/assets/content/blog/charbon-podcast-with-claire.png",[925,1239],"french","2023-04-26",{"title":1583,"path":1584,"stem":1585,"meta":1586},"Podcast speaker on SuperDataScience Podcast #669","/framework/blog/superdatascience-with-adrian","framework/blog/922.superdatascience-with-adrian",{"layout":925,"thumbnail":1587,"tags":1589,"date":1590,"hidden":931},{"src":1588},"https://img.youtube.com/vi/AjxLnll8QHg/maxresdefault.jpg",[925,1239],"2023-04-11",{"title":1592,"path":1593,"stem":1594,"meta":1595},"Pathway featured in Sifted briefing on GenAI","/framework/blog/pathway-featured-in-sifted-briefing-on-genai","framework/blog/923.pathway-featured-in-sifted-briefing-on-genai",{"layout":925,"thumbnail":1596,"tags":1598,"date":1599,"hidden":931},{"src":1597},"/assets/content/blog/sifted-th.png",[925,1112],"2023-03-09",{"title":1601,"path":1602,"stem":1603,"meta":1604},"Improving asset utilization with Pathway: combining IoT data with real-time data processing","/framework/blog/improving-asset-utilization","framework/blog/924.improving-asset-utilization",{"layout":925,"thumbnail":1605,"tags":1607,"date":1608,"hidden":931},{"src":1606},"/assets/content/blog/improving-asset-utilzation-th.png",[925],"2023-04-06",{"title":1610,"path":1611,"stem":1612,"meta":1613},"What is Fuzzy Join and How Can it Help You Make Sense of Your Data?","/framework/blog/fuzzy-join","framework/blog/925.fuzzy-join",{"layout":925,"thumbnail":1614,"tags":1616,"date":1617,"hidden":931},{"src":1615},"/assets/content/blog/fuzzy-join/th-fuzzy-join.svg",[925,1066,1006],"2023-04-05",{"title":1619,"path":1620,"stem":1621,"meta":1622},"Discussing supply chain analytics on the Data Engineering Podcast","/framework/blog/data-engineering-podcast","framework/blog/926.data-engineering-podcast",{"layout":925,"thumbnail":1623,"tags":1625,"date":1626,"hidden":931},{"src":1624},"/assets/content/blog/data-engineering-podcast.jpg",[925,1239],"2022-10-16",{"title":1628,"path":1629,"stem":1630,"meta":1631},"The value of real-time analytics - IoT for all podcast","/framework/blog/iot-for-all-podcast","framework/blog/927.Iot-for-all-podcast",{"layout":925,"thumbnail":1632,"tags":1634,"date":1635,"hidden":931},{"src":1633},"https://i3.ytimg.com/vi/4TSK0fpMC4I/maxresdefault.jpg",[925,1239],"2023-02-20",{"title":1637,"path":1638,"stem":1639,"meta":1640},"Guest speaker at Maddyness Keynote","/framework/blog/maddy-keynote-2023","framework/blog/928.Maddy-keynote-2023",{"layout":925,"thumbnail":1641,"tags":1643,"date":1644,"hidden":931},{"src":1642},"/assets/content/blog/maddy-keynote.jpg",[925,1112],"2023-02-08",{"title":1646,"path":1647,"stem":1648,"meta":1649},"Guest speakers on Female Foundry Podcast","/framework/blog/female-foundry-podcast","framework/blog/929.Female-Foundry-Podcast",{"layout":925,"thumbnail":1650,"tags":1652,"date":1653,"hidden":931},{"src":1651},"/assets/content/blog/female-foundry-podcast.jpg",[925,1239],"2023-02-11",{"title":1655,"path":1656,"stem":1657,"meta":1658},"Podcast speaker on Super Data Science Podcast","/framework/blog/liquidneuralnetworks","framework/blog/930.LiquidNeuralNetworks",{"layout":925,"date":1659,"thumbnail":1660,"tags":1662,"hidden":931},"2022-12-02",{"src":1661},"/assets/content/blog/LiquidNeurealnetwork.jpg",[925,1238,1239],{"title":1664,"path":1665,"stem":1666,"meta":1667},"Pathway Premieres at PyData Global","/framework/blog/pydata","framework/blog/932.PyData",{"layout":925,"date":1668,"thumbnail":1669,"tags":1671,"hidden":931},"2022-12-01",{"src":1670,"contain":931},"/assets/content/blog/logo-conference-pydata-global-2021.png",[925],{"title":1673,"path":1674,"stem":1675,"meta":1676},"Pathway at ODSC West","/framework/blog/odsc","framework/blog/934.ODSC",{"layout":925,"date":1677,"thumbnail":1678,"tags":1680,"hidden":931},"2022-11-01",{"src":1679},"/assets/content/blog/ODSCWest.png",[925,1681],"culture",{"title":1683,"path":1684,"stem":1685,"meta":1686},"Guest speaker on IoT For All Podcast","/framework/blog/iotforall","framework/blog/936.IoTforAll",{"layout":925,"thumbnail":1687,"tags":1689,"date":1690,"hidden":931},{"src":1688},"/assets/content/blog/IoT-for-all.png",[925,1238,1239],"2022-08-01",{"title":1692,"path":1693,"stem":1694,"meta":1695},"Pathway graduated from CDL-Montreal!","/framework/blog/cdl","framework/blog/938.CDL",{"layout":925,"date":1696,"thumbnail":1697,"tags":1699,"hidden":931},"2022-06-01",{"src":1698},"/assets/content/blog/Pathway-CDL-Montreal.jpeg",[925],{"title":1701,"path":1702,"stem":1703,"meta":1704},"Pathway helps La Poste reduce IoT costs by 50%","/framework/blog/laposte","framework/blog/940.LaPoste",{"layout":925,"date":1696,"thumbnail":1705,"tags":1707,"hidden":931},{"src":1706,"contain":931,"padding":931},"/assets/content/blog/ROILaposte.png",[925],{"title":1709,"path":1710,"stem":1711,"meta":1712},"Why Pathway","/framework/blog/why-pathway","framework/blog/945.why-pathway",{"layout":925,"date":1713,"thumbnail":1714,"tags":1716,"hidden":931},"2022-05-24",{"src":1715},"/assets/content/blog/Pathwaybanner.jpg",[925],{"title":1718,"path":1719,"stem":1720,"meta":1721},"Pathway wins the #Spring50 Pitch Contest!","/framework/blog/spring50","framework/blog/950.spring50",{"layout":925,"date":1722,"thumbnail":1723,"tags":1725,"hidden":931},"2022-05-13",{"src":1724,"provider":10,"contain":931},"/assets/blog/thumbnails/digital-50-contest.png",[925],{"title":1727,"path":1728,"stem":1729,"meta":1730},"Pathway is a WomenTech EU Laureate","/framework/blog/womentecheu","framework/blog/960.WomenTechEU",{"layout":925,"date":1731,"thumbnail":1732,"tags":1734,"hidden":931},"2022-03-01",{"src":1733,"provider":10},"/assets/blog/thumbnails/co-funded-th.png",[925],{"title":1736,"path":1737,"stem":1738,"meta":1739},"Pathway is a Gartner Representative Vendor","/framework/blog/gartner","framework/blog/970.Gartner",{"layout":925,"date":1740,"thumbnail":1741,"tags":1743,"enterprise":931,"hidden":931},"2021-10-19",{"src":1742,"provider":10,"contain":931},"/assets/blog/thumbnails/gartner-th.png",[925],{"title":1745,"path":1746,"stem":1747,"meta":1748},"Pathway joins Agoranov, French Science and Tech incubator - in Paris, France","/framework/blog/agoranov","framework/blog/980.Agoranov",{"layout":925,"date":1749,"thumbnail":1750,"tags":1752,"hidden":931},"2021-10-09",{"src":1751,"provider":10,"contain":931},"/assets/blog/thumbnails/Agoranov-Logo.png",[925],{"title":1754,"path":1755,"stem":1756,"meta":1757},"Pathway has been selected by Hello Tomorrow as a Deep Tech Pioneer ","/framework/blog/hellotomorrow","framework/blog/990.hellotomorrow",{"layout":925,"date":1758,"thumbnail":1759,"tags":1761,"hidden":931},"2021-09-27",{"src":1760,"provider":10,"contain":931},"/assets/blog/thumbnails/hello-tomorrow-th.png",[925],{"title":1763,"path":1764,"stem":1765,"meta":1766},"Gartner® recognizes Pathway as an Emerging Visionary in GenAI Engineering","/framework/blog/gartner-gen-ai-engineering","framework/blog/996.gartner-gen-ai-engineering",{"layout":925,"date":1767,"thumbnail":1768,"tags":1770,"related":1771},"2024-11-14",{"src":1769,"provider":10},"/assets/blog/thumbnails/th-pathway-gartner-emerging-market-quadrants.png",[1112],[1309],{"title":1773,"path":1774,"stem":1775,"meta":1776},"Adaptive Agents for Real-Time RAG: Domain-Specific AI for Legal, Finance & Healthcare","/framework/blog/adaptive-agents-rag","framework/blog/adaptive-agents-rag",{"layout":925,"date":1777,"thumbnail":1778,"tags":1780},"2025-04-02",{"src":1779,"provider":10},"/assets/blog/thumbnails/adaptive-agents-rag-th.png",[930],{"title":1782,"path":1783,"stem":1784,"meta":1785},"How AI Agents in Finance Are Transforming Financial Due Diligence: FA3STER","/framework/blog/ai-agents-finance-due-diligence","framework/blog/ai-agents-finance-due-diligence",{"layout":925,"date":1786,"thumbnail":1787,"tags":1789},"2025-03-26",{"src":1788,"provider":10},"/assets/blog/thumbnails/ai-agents-finance-due-diligence-th.png",[930],{"title":5,"path":932,"stem":935,"meta":1791},{"layout":925,"date":926,"thumbnail":1792,"tags":1793},{"src":928,"provider":10},[930],{"title":1795,"path":1796,"stem":1797,"meta":1798},"LiveAI™ for SEC Filings Analysis","/framework/blog/ai-for-sec-filings","framework/blog/ai-for-sec-filings",{"layout":925,"date":1799,"thumbnail":1800,"tags":1802},"2025-07-09",{"src":1801,"provider":10},"/assets/blog/thumbnails/ai-for-sec-filings-th.png",[930],{"title":1804,"path":1805,"stem":1806,"meta":1807},"AI Paper Reviewer | LiveAI™ for Conference Classification","/framework/blog/ai-paper-reviewer","framework/blog/ai-paper-reviewer",{"layout":925,"date":1808,"thumbnail":1809,"tags":1811},"2025-06-19",{"src":1810,"provider":10},"/assets/blog/thumbnails/ai-paper-reviewer-th.png",[930],{"title":1813,"path":1814,"stem":1815,"meta":1816},"LiveAI™ Tools for Equity Research & Compliance Management","/framework/blog/ai-tools-for-equity-analysis","framework/blog/ai-tools-for-equity-analysis",{"layout":925,"date":1817,"thumbnail":1818,"tags":1820},"2025-04-23",{"src":1819,"provider":10},"/assets/blog/thumbnails/live-ai-legal-compliance-financial-analysis-th.png",[930],{"title":1822,"path":1823,"stem":1824,"meta":1825},"Pathway’s BDH solves Sudoku Extreme with 97.4% accuracy, while leading LLMs are close to 0","/framework/blog/beyond-transformers-sudoku-bench","framework/blog/beyond-transformers-sudoku-bench",{"date":1826,"thumbnail":1827,"tags":1829,"related":951,"redirect":1831},"2026-03-17",{"src":1828,"provider":10},"/assets/blog/thumbnails/the-sudoku-test-th.png",[1830],"bdh","/research/beyond-transformers-sudoku-bench",{"title":1833,"path":1834,"stem":1835,"meta":1836},"Build RAG Apps in YAML - Recording from the Intel AI DevSummit","/framework/blog/build-rag-apps-in-yaml-recording-from-the-intel-ai-devsummit","framework/blog/build-rag-apps-in-yaml-recording-from-the-intel-ai-devsummit",{"aside":951,"layout":925,"date":1051,"thumbnail":1837,"tags":1839,"hidden":931},{"src":1838},"https://i3.ytimg.com/vi/OEEZ5qSHJSs/hqdefault.jpg",[1238,1066,1006],{"title":1841,"path":1842,"stem":1843,"meta":1844},"Data Discovery to Data Pipeline Process","/framework/blog/data-discovery-to-data-pipeline-process","framework/blog/data-discovery-to-data-pipeline-process",{"layout":925,"date":1845,"thumbnail":1846,"tags":1848},"2025-02-28",{"src":1847,"provider":10},"/assets/blog/thumbnails/data-discovery-th.png",[930],{"title":1850,"path":1851,"stem":1852,"meta":1853},"Exploring Kappa Architecture with Pathway","/framework/blog/exploring-kappa-architecture-with-pathway","framework/blog/exploring-kappa-architecture-with-pathway",{"single":931,"aside":951,"layout":925,"date":1845,"thumbnail":1854,"tags":1856},{"src":1855,"provider":10},"/assets/blog/thumbnails/kappa-architecture-th.png",[930],{"title":1858,"path":1859,"stem":1860,"meta":1861},"Resilient Financial API Integration with LiveAI™","/framework/blog/financial-api-integration-live-ai","framework/blog/financial-api-integration-live-ai",{"layout":925,"date":1862,"thumbnail":1863,"tags":1865},"2025-05-23",{"src":1864,"provider":10},"/assets/blog/thumbnails/financial-api-integration-live-ai-th.png",[930],{"title":1867,"path":1868,"stem":1869,"meta":1870},"LiveAI™ for Financial Intelligence with Event-Based State Machine","/framework/blog/financial-intelligence-with-event-based-state-machine","framework/blog/financial-intelligence-with-event-based-state-machine",{"keywords":1871,"layout":925,"date":1879,"thumbnail":1880,"tags":1882},[1872,1873,1874,1875,1876,1877,1878],"Large vector stores","Event-Based State Machine","dynamic data sources","reliability","hallucination","Financial Intelligence","LiveAI™ framework","2025-06-12",{"src":1881,"provider":10},"/assets/blog/thumbnails/financial-intelligence-with-event-based-state-machine-th.png",[930],{"title":1884,"path":1885,"stem":1886,"meta":1887},"Building a Real-Time Radiology AI System with Pathway and LandingAI","/framework/blog/landing-ai","framework/blog/landing-ai",{"layout":925,"date":1888,"thumbnail":1889,"tags":1891},"2025-10-02",{"src":1890,"provider":10},"/assets/blog/thumbnails/landing-ai-th.png",[930],{"title":1893,"path":1894,"stem":1895,"meta":1896},"Pathway MCP Server: Live Indexing & Analytics for your Agents","/framework/blog/live-ai-mcp-server","framework/blog/live-ai-mcp-server",{"aside":951,"layout":925,"date":1897,"thumbnail":1898,"tags":1900,"coauthors":1901,"hidden":931},"2025-08-22",{"src":1899,"contain":931},"/assets/content/blog/MCP/mcp-server-thumbnail.png",[1066,1006],[1902],{"name":1903,"description\"":1904,"img":1905,"provider":10,"linkedin":1906},"Mudit Srivastava","Director of Growth","/assets/authors/mudit-srivastava.jpg","https://www.linkedin.com/in/muditjps/",{"title":1908,"path":1909,"stem":1910,"meta":1911},"Multi Agentic RAG & LiveAI™ for Finance and Legal Solutions","/framework/blog/live-ai-multi-agentic-rag","framework/blog/live-ai-multi-agentic-rag",{"layout":925,"date":1912,"thumbnail":1913,"tags":1915},"2025-04-09",{"src":1914,"provider":10},"/assets/blog/thumbnails/live-ai-multi-agentic-rag-th.png",[930],{"title":1917,"path":1918,"stem":1919,"meta":1920},"Multi Agent RAG with Interleaved Retrieval and Reasoning for Long Docs","/framework/blog/multi-agent-rag-interleaved-retrieval-reasoning","framework/blog/multi-agent-rag-interleaved-retrieval-reasoning",{"layout":925,"date":1921,"thumbnail":1922,"tags":1924},"2025-03-19",{"src":1923,"provider":10},"/assets/blog/thumbnails/multi-agent-rag-system-for-long-document-question-answering-th.png",[930],{"title":1926,"path":1927,"stem":1928,"meta":1929},"Real-Time Multimodal Data Processing with Pathway and Docling","/framework/blog/multimodal-data-processing","framework/blog/multimodal-data-processing",{"aside":951,"layout":925,"date":1930,"thumbnail":1931,"tags":1933,"coauthors":1934,"hidden":931},"2025-05-30",{"src":1932,"contain":931},"/assets/content/blog/docling-parser/docling-th.png",[1066,1006],[1935],{"name":1936,"description\"":1937,"img":1938,"linkedin":1939},"Albert Roethel","AI Engineer","/assets/content/blog/avatars/albert.jpg","https://www.linkedin.com/in/albertroethel/",{"title":1941,"path":1942,"stem":1943,"meta":1944},"Regulatory Compliance Automation with LiveAI™","/framework/blog/regulatory-compliance-automation-live-ai","framework/blog/regulatory-compliance-automation-live-ai",{"layout":925,"date":1945,"thumbnail":1946,"tags":1948,"related":951},"2025-04-30",{"src":1947,"provider":10},"/assets/blog/thumbnails/ai-legal-document-analysis-th.png",[930],{"title":1950,"path":1951,"stem":1952,"meta":1953},"Unlocking data stream processing [Part 1] - real-time linear regression","/framework/blog/unlocking-data-stream-processing-1","framework/blog/unlocking-data-stream-processing-1",{"layout":925,"redirection":931,"date":1954,"thumbnail":1955,"tags":1957,"hidden":931},"2023-02-16",{"src":1956},"/assets/content/tutorials/unlocking-data-stream-processing-1-th.png",[1066,1006],{"title":1959,"path":1960,"stem":1961,"meta":1962},"Unlocking data stream processing [Part 2] - realtime server logs monitoring with a sliding window","/framework/blog/unlocking-data-stream-processing-2","framework/blog/unlocking-data-stream-processing-2",{"layout":925,"redirection":931,"date":1599,"thumbnail":1963,"tags":1965,"hidden":931},{"src":1964},"/assets/content/tutorials/unlocking-data-stream-processing-2-th.png",[1066,1006],{"title":1967,"path":1968,"stem":1969,"meta":1970},"Unlocking data stream processing [Part 3] - data enrichment with fuzzy joins","/framework/blog/unlocking-data-stream-processing-3","framework/blog/unlocking-data-stream-processing-3",{"layout":925,"redirection":931,"date":1971,"thumbnail":1972,"tags":1974,"hidden":931},"2023-05-09",{"src":1973},"/assets/content/tutorials/unlocking-data-stream-processing-3-th.png",[1066,1006],{"layout":950,"aside":951,"toc":951,"single":931},{"id":1977,"title":5,"author":1978,"body":1979,"date":2594,"description":922,"extension":923,"hidden":951,"keywords":2595,"meta":2596,"navigation":931,"path":932,"seo":2597,"stem":935,"tags":2598,"thumbnail":2599,"__hash__":2600},"blog/framework/blog/ai-financial-report-analysis.md",{"name":7,"description":8,"img":9,"provider":10},{"type":12,"value":1980,"toc":2565},[1981,1983,1987,1989,1991,1995,2007,2009,2027,2029,2039,2041,2043,2045,2047,2073,2075,2077,2079,2081,2083,2085,2087,2089,2092,2094,2096,2098,2100,2102,2105,2127,2129,2131,2153,2155,2157,2159,2161,2163,2165,2167,2169,2171,2175,2179,2183,2187,2191,2193,2195,2199,2203,2205,2209,2215,2217,2219,2223,2227,2229,2231,2233,2251,2253,2255,2257,2259,2277,2281,2283,2341,2346,2374,2379,2381,2383,2385,2387,2391,2393,2397,2399,2403,2405,2409,2411,2433,2435,2437,2439,2459,2461,2521,2523],[15,1982,18],{"id":17},[20,1984,22,1985],{},[24,1986,26],{},[20,1988,29],{},[20,1990,32],{},[20,1992,1993,39],{},[36,1994,38],{},[20,1996,42,1997,46,1999,50,2001,54,2003,58,2005,62],{},[36,1998,45],{},[36,2000,49],{},[36,2002,53],{},[36,2004,57],{},[36,2006,61],{},[64,2008,67],{"id":66},[69,2010,2011,2015,2019,2023],{},[72,2012,2013,77],{},[36,2014,76],{},[72,2016,2017,83],{},[36,2018,82],{},[72,2020,2021,89],{},[36,2022,88],{},[72,2024,2025,95],{},[36,2026,94],{},[64,2028,99],{"id":98},[69,2030,2031,2033,2035,2037],{},[72,2032,104],{},[72,2034,107],{},[72,2036,110],{},[72,2038,113],{},[20,2040,116],{},[15,2042,120],{"id":119},[64,2044,124],{"id":123},[20,2046,127],{},[69,2048,2049,2053,2057,2061,2065,2069],{},[72,2050,2051,135],{},[36,2052,134],{},[72,2054,2055,141],{},[36,2056,140],{},[72,2058,2059,147],{},[36,2060,146],{},[72,2062,2063,153],{},[36,2064,152],{},[72,2066,2067,159],{},[36,2068,158],{},[72,2070,2071,165],{},[36,2072,164],{},[64,2074,169],{"id":168},[171,2076],{"src":173},[64,2078,177],{"id":176},[179,2080],{"href":181,"title":182,"website":183},[64,2082,187],{"id":186},[20,2084,190],{},[20,2086,193],{},[20,2088,196],{},[198,2090],{":zoomable":200,"alt":201,"className":2091,"provider":10,"src":206},[203,204,205],[64,2093,210],{"id":209},[20,2095,213],{},[20,2097,216],{},[64,2099,220],{"id":219},[20,2101,223],{},[198,2103],{":zoomable":200,"alt":201,"className":2104,"provider":10,"src":227},[203,204,205],[69,2106,2107,2111,2115,2119,2123],{},[72,2108,2109,235],{},[36,2110,234],{},[72,2112,2113,241],{},[36,2114,240],{},[72,2116,2117,247],{},[36,2118,246],{},[72,2120,2121,253],{},[36,2122,252],{},[72,2124,2125,259],{},[36,2126,258],{},[64,2128,263],{"id":262},[20,2130,266],{},[268,2132,2133,2137,2141,2145,2149],{},[72,2134,2135,275],{},[36,2136,274],{},[72,2138,2139,281],{},[36,2140,280],{},[72,2142,2143,287],{},[36,2144,286],{},[72,2146,2147,293],{},[36,2148,292],{},[72,2150,2151,299],{},[36,2152,298],{},[64,2154,303],{"id":302},[20,2156,306],{},[15,2158,310],{"id":309},[64,2160,314],{"id":313},[20,2162,317],{},[20,2164,320],{},[64,2166,324],{"id":323},[20,2168,327],{},[64,2170,331],{"id":330},[20,2172,2173,337],{},[36,2174,336],{},[20,2176,2177,343],{},[36,2178,342],{},[20,2180,2181,349],{},[36,2182,348],{},[20,2184,2185],{},[36,2186,354],{},[69,2188,2189],{},[72,2190,359],{},[64,2192,363],{"id":362},[20,2194,366],{},[20,2196,2197],{},[36,2198,371],{},[373,2200,2201],{},[20,2202,377],{},[20,2204,380],{},[373,2206,2207],{},[20,2208,385],{},[373,2210,2211],{},[20,2212,390,2213,395],{},[392,2214,394],{},[20,2216,398],{},[20,2218,401],{},[373,2220,2221],{},[20,2222,406],{},[373,2224,2225],{},[20,2226,411],{},[20,2228,414],{},[64,2230,418],{"id":417},[20,2232,421],{},[69,2234,2235,2239,2243,2247],{},[72,2236,2237,429],{},[36,2238,428],{},[72,2240,2241,435],{},[36,2242,434],{},[72,2244,2245,441],{},[36,2246,440],{},[72,2248,2249,447],{},[36,2250,446],{},[20,2252,450],{},[64,2254,454],{"id":453},[20,2256,457],{},[20,2258,460],{},[69,2260,2261,2265,2269,2273],{},[72,2262,2263,467],{},[36,2264,274],{},[72,2266,2267,472],{},[36,2268,280],{},[72,2270,2271,477],{},[36,2272,286],{},[72,2274,2275,482],{},[36,2276,292],{},[20,2278,2279],{},[36,2280,487],{},[20,2282,490],{},[492,2284,498,2286,498,2318],{"className":2285},[495,203,496,497],[500,2287,502,2288,502,2300,498],{},[504,2289,506,2290,506,2292,506,2294,506,2296,506,2298,502],{},[508,2291,511],{"rowSpan":510},[508,2293,515],{"colSpan":514},[508,2295,518],{"colSpan":514},[508,2297,521],{"colSpan":514},[508,2299,524],{"colSpan":514},[504,2301,506,2302,2304,506,2306,2308,506,2310,2312,506,2314,2316,502],{},[508,2303,529],{},[508,2305,532],{},[508,2307,529],{},[508,2309,532],{},[508,2311,529],{},[508,2313,532],{},[508,2315,529],{},[508,2317,532],{},[546,2319,502,2320,498],{},[504,2321,506,2322,506,2325,2327,506,2329,2331,506,2333,2335,506,2337,2339,502],{},[551,2323,555],{"className":2324,"rowSpan":510},[554],[551,2326,558],{},[551,2328,561],{},[551,2330,564],{},[551,2332,567],{},[551,2334,570],{},[551,2336,573],{},[551,2338,576],{},[551,2340,579],{},[20,2342,2343],{},[24,2344,587],{"className":2345},[585,203,497,586],[492,2347,498,2349,498,2363],{"className":2348},[495,203,496,497],[500,2350,502,2351,502,2357,498],{},[504,2352,506,2353,506,2355,502],{},[508,2354,511],{"rowSpan":510},[508,2356,599],{"colSpan":514},[504,2358,506,2359,2361,502],{},[508,2360,529],{},[508,2362,532],{},[546,2364,502,2365,498],{},[504,2366,506,2367,506,2370,2372,502],{},[551,2368,555],{"className":2369,"rowSpan":510},[554],[551,2371,615],{},[551,2373,618],{},[20,2375,2376],{},[24,2377,624],{"className":2378},[585,203,497,586],[64,2380,298],{"id":61},[20,2382,629],{},[15,2384,633],{"id":632},[20,2386,636],{},[20,2388,2389],{},[36,2390,641],{},[20,2392,644],{},[20,2394,2395],{},[36,2396,649],{},[20,2398,652],{},[20,2400,2401],{},[36,2402,657],{},[20,2404,660],{},[20,2406,2407],{},[36,2408,665],{},[20,2410,668],{},[69,2412,2413,2417,2421,2425,2429],{},[72,2414,673,2415,677],{},[36,2416,676],{},[72,2418,680,2419,677],{},[36,2420,683],{},[72,2422,686,2423,677],{},[36,2424,689],{},[72,2426,692,2427,677],{},[36,2428,695],{},[72,2430,698,2431,677],{},[36,2432,701],{},[15,2434,705],{"id":704},[20,2436,708],{},[20,2438,711],{},[69,2440,2441,2445,2449,2454],{},[72,2442,2443],{},[717,2444,720],{"href":719},[72,2446,2447],{},[717,2448,726],{"href":725},[72,2450,2451],{},[717,2452,734],{"href":731,"rel":2453},[733],[72,2455,2456],{},[717,2457,741],{"href":739,"rel":2458},[733],[15,2460,745],{"id":744},[747,2462,2463,2467],{"type":749},[751,2464,2465],{"v-slot:summary":201},[20,2466,755],{},[751,2468,2469],{"v-slot:content":201},[268,2470,2471,2476,2481,2486,2491,2496,2501,2506,2511,2516],{},[72,2472,2473],{},[717,2474,766],{"href":764,"rel":2475},[733],[72,2477,2478],{},[717,2479,773],{"href":771,"rel":2480},[733],[72,2482,2483],{},[717,2484,780],{"href":778,"rel":2485},[733],[72,2487,2488],{},[717,2489,787],{"href":785,"rel":2490},[733],[72,2492,2493],{},[717,2494,794],{"href":792,"rel":2495},[733],[72,2497,2498],{},[717,2499,801],{"href":799,"rel":2500},[733],[72,2502,2503],{},[717,2504,808],{"href":806,"rel":2505},[733],[72,2507,2508],{},[717,2509,815],{"href":813,"rel":2510},[733],[72,2512,2513],{},[717,2514,822],{"href":820,"rel":2515},[733],[72,2517,2518],{},[717,2519,829],{"href":827,"rel":2520},[733],[15,2522,833],{"id":832},[69,2524,2525,2530,2535,2540,2545,2550,2555,2560],{},[72,2526,2527],{},[717,2528,842],{"href":840,"rel":2529},[733],[72,2531,2532],{},[717,2533,849],{"href":847,"rel":2534},[733],[72,2536,2537],{},[717,2538,856],{"href":854,"rel":2539},[733],[72,2541,2542],{},[717,2543,863],{"href":861,"rel":2544},[733],[72,2546,2547],{},[717,2548,870],{"href":868,"rel":2549},[733],[72,2551,2552],{},[717,2553,877],{"href":875,"rel":2554},[733],[72,2556,2557],{},[717,2558,884],{"href":882,"rel":2559},[733],[72,2561,2562],{},[717,2563,891],{"href":889,"rel":2564},[733],{"title":201,"searchDepth":510,"depth":510,"links":2566},[2567,2571,2581,2590,2591,2592,2593],{"id":17,"depth":510,"text":18,"children":2568},[2569,2570],{"id":66,"depth":897,"text":67},{"id":98,"depth":897,"text":99},{"id":119,"depth":510,"text":120,"children":2572},[2573,2574,2575,2576,2577,2578,2579,2580],{"id":123,"depth":897,"text":124},{"id":168,"depth":897,"text":169},{"id":176,"depth":897,"text":177},{"id":186,"depth":897,"text":187},{"id":209,"depth":897,"text":210},{"id":219,"depth":897,"text":220},{"id":262,"depth":897,"text":263},{"id":302,"depth":897,"text":303},{"id":309,"depth":510,"text":310,"children":2582},[2583,2584,2585,2586,2587,2588,2589],{"id":313,"depth":897,"text":314},{"id":323,"depth":897,"text":324},{"id":330,"depth":897,"text":331},{"id":362,"depth":897,"text":363},{"id":417,"depth":897,"text":418},{"id":453,"depth":897,"text":454},{"id":61,"depth":897,"text":298},{"id":632,"depth":510,"text":633},{"id":704,"depth":510,"text":705},{"id":744,"depth":510,"text":745},{"id":832,"depth":510,"text":833},"2025-09-03T00:00:00.000Z",null,{"layout":925},{"title":5,"description":922},[930],{"src":928,"provider":10},"2buKD2Mv9JpozA99llddtg1_ko8v4oGZI8iWKUB2Yic",[2602,2640,2776,2808,2840,2874,2904,2935,2962,2990,3021,3053,3085,3116,3170,3216,3301,3331,3363,3395,3424,3455,3484,3544,3621,3652,3725,3756,3785,3817,3846,3879,3912,3945,3974,4045,4073,4104,4148,4179,4236,4283,4315,4345,4377,4408,4457,4488,4518,4549,4579,4610,4642,4673,4701,4747,4775,4805,4833,4864,4895,4925,4956,4987,5018,5046,5074,5097,5119,5142,5165,5188,5210,5232,5254,5284,5315,5346,5396,5426,5455,5485,5515,5545,5576,5608,5637,5668,5696,5725,5755,5784,5886,5917,5948,5977,6019,6175,6203,8152],{"id":2603,"title":2604,"author":2605,"body":2609,"description":2629,"extension":923,"meta":2630,"navigation":931,"path":2635,"seo":2636,"sitemap":2637,"stem":2638,"__hash__":2639},"content/9.news/100-women-in-tech-2025.md","100 Women in Tech",{"name":2606,"description":201,"website":2607,"img":2608,"provider":10},"sifted.eu","https://sifted.eu/","/assets/blog/avatars/sifted-av.png",{"type":12,"value":2610,"toc":2627},[2611,2616,2624],[2612,2613,2615],"h1",{"id":2614},"taking-you-to-an-external-site","Taking you to an external site",[20,2617,2618,2619,2623],{},"You will be taken to ",[717,2620,2621],{"href":2621,"rel":2622},"https://sifted.eu/list/100-women-in-tech-2025",[733]," in a moment.",[2625,2626],"redirect",{"url":2621},{"title":201,"searchDepth":510,"depth":510,"links":2628},[],"Spotlighting 100 women who are making a real impact and shaping the future of Europe's tech ecosystem",{"layout":925,"redirection":931,"thumbnail":2631,"tags":2633,"date":2634},{"src":2632,"contain":931},"https://www.datocms-assets.com/60124/1758796746-copy-of-nominate-a-female-rising-star-1.png",[1112],"2025-10-10","/news/100-women-in-tech-2025",{"title":2604,"description":2629},{"loc":2635},"9.news/100-women-in-tech-2025","75jFMS8DWTMmQo3XArzmGjst8CZNtgzml0f8jLPlOoY",{"id":2641,"title":2642,"author":2643,"body":2650,"description":2765,"extension":923,"meta":2766,"navigation":931,"path":2771,"seo":2772,"sitemap":2773,"stem":2774,"__hash__":2775},"content/9.news/845.la-poste-optimizes-colissimo-flows-in-real-time.md","La Poste Optimizes Colissimo Flows in Real Time - Modern Data Stack Recording available",{"id":2644,"url":2645,"name":2646,"description":2647,"img":2648,"provider":10,"linkedin":2649},"claire","claire-nouet","Claire Nouet","COO","/assets/authors/claire-nouet.jpg","https://www.linkedin.com/in/clairenouet/",{"type":12,"value":2651,"toc":2760},[2652,2655,2664,2667,2676,2680,2694,2698,2718,2722],[20,2653,2654],{},"La Poste Group is using Pathway’s real-time data processing capabilities to optimize the flows of its well-known Colissimo business line. Jean-Paul Fabre, the Head of Technological innovation at La Poste, and Claire Nouet, the co-founder of Pathway, discussed the collaboration during the Modern Data Stack summit in Paris.",[20,2656,2657,2658,2663],{},"Learn how ",[717,2659,2662],{"href":2660,"rel":2661},"https://www.linkedin.com/company/la-poste-groupe/",[733],"La Poste Group"," deployed ‘operational speed’ AI to forecast more accurately, assess disruption and automate key processes, improving operations in a meaningful way! (The video is in French, recap below!)",[171,2665],{"src":2666},"https://www.youtube.com/watch?v=RyZFWeADJXM",[2668,2669,2673],"pathway-button",{"className":2670,"href":2671,"icon":2672},[204],"/success-stories/la-poste","heroicons:check-badge-20-solid",[20,2674,2675],{},"Read more about the case study",[15,2677,2679],{"id":2678},"challenges-faced-by-la-poste","Challenges Faced by La Poste",[69,2681,2682,2685,2688,2691],{},[72,2683,2684],{},"La Poste handles a high volume of packages across 17 industrial platforms, has more than 400 truck movements daily, and 16 million+ unused data points.",[72,2686,2687],{},"Need to provide platform operators with real-time information on truck arrival times, origins, and destinations.",[72,2689,2690],{},"Requirement to improve efficiency to avoid congestion and incidents.",[72,2692,2693],{},"Need to identify anomalies in real-time",[15,2695,2697],{"id":2696},"key-objectives-and-benefits","Key Objectives and Benefits",[69,2699,2700,2706,2712],{},[72,2701,2702,2705],{},[36,2703,2704],{},"Reduce costs",": Pathway's solution is more cost-effective than existing outsourced solutions, with savings reinvested in functional improvements.",[72,2707,2708,2711],{},[36,2709,2710],{},"Leverage data",": La Poste generates approximately 16 million geolocation points annually but was not fully utilizing this data. Pathway helps make sense of this data and turn it into actionable insights.",[72,2713,2714,2717],{},[36,2715,2716],{},"Simplify infrastructure",": Pathway enables easier integration of different acquisition platforms and sensors and facilitates predictive calculations and rapid prototyping.",[15,2719,2721],{"id":2720},"how-pathway-works","How Pathway Works",[69,2723,2724,2730,2736,2742,2748,2754],{},[72,2725,2726,2729],{},[36,2727,2728],{},"Network Identification",": Pathway identifies nodes (e.g. locations where trucks stop for a significant time) within the network. It differentiates between relevant nodes (e.g., platforms) and irrelevant ones (e.g., driver rest stops).",[72,2731,2732,2735],{},[36,2733,2734],{},"Route Analysis",": Pathway determines the primary routes between platforms and identifies alternative routes. This helps La Poste understand if drivers are using preferred (e.g., tolled) routes or opting for alternative routes.",[72,2737,2738,2741],{},[36,2739,2740],{},"Data Integration",": Pathway concentrates real-time geolocation data and historical data on a single platform. This creates a digital twin of the network, which can be used for real-time monitoring and analysis.",[72,2743,2744,2747],{},[36,2745,2746],{},"Real-time data processing",": Data scientists can work with real-time data in Jupiter Notebooks, and code can be moved directly into production, improving productivity.",[72,2749,2750,2753],{},[36,2751,2752],{},"Anomaly Detection",": Uses machine learning to detect anomalies with security implications.",[72,2755,2756,2759],{},[36,2757,2758],{},"GPS Data Enhancement",": Pathway automatically creates polygons based on GPS quality to filter out errors and false positives caused by signal fluctuations and metallic buildings.",{"title":201,"searchDepth":510,"depth":510,"links":2761},[2762,2763,2764],{"id":2678,"depth":510,"text":2679},{"id":2696,"depth":510,"text":2697},{"id":2720,"depth":510,"text":2721},"La Poste Group is using Pathway’s real-time data processing capabilities to optimize the flows of its well-known Colissimo business line. Jean-Paul Fabre, the Head of Technological innovation at La Poste, and Claire Nouet, the co-founder of Pathway, discussed the collaboration during the Modern Data Stack summit in Paris",{"layout":925,"thumbnail":2767,"tags":2769,"date":2770,"hidden":931},{"src":2768},"https://i3.ytimg.com/vi/RyZFWeADJXM/maxresdefault.jpg",[1112,1238],"2025-03-04","/news/la-poste-optimizes-colissimo-flows-in-real-time",{"title":2642,"description":2765},{"loc":2771},"9.news/845.la-poste-optimizes-colissimo-flows-in-real-time","ZlgD42RrarQpX3dA_P_POvRxiJ72MiTbnv8SbO0keyc",{"id":2777,"title":2778,"author":2779,"body":2783,"description":2797,"extension":923,"meta":2798,"navigation":931,"path":2803,"seo":2804,"sitemap":2805,"stem":2806,"__hash__":2807},"content/9.news/846.becoming-ai-savvy-for-transformation.md","Becoming AI-savvy: going beyond data smarts for business transformation",{"name":2780,"img":2781,"website":2782},"TechInformed","https://www.google.com/s2/favicons?domain=techinformed.com&sz=128","https://techinformed.com",{"type":12,"value":2784,"toc":2795},[2785,2787,2793],[2612,2786,2615],{"id":2614},[20,2788,2618,2789,2623],{},[717,2790,2791],{"href":2791,"rel":2792},"https://techinformed.com/becoming-ai-savvy-for-transformation/",[733],[2625,2794],{"url":2791},{"title":201,"searchDepth":510,"depth":510,"links":2796},[],"Claire Nouet, COO and Co-founder of Pathway, outlines how business leaders need to look beyond data smarts to become AI-savvy",{"layout":925,"redirection":931,"thumbnail":2799,"tags":2801,"date":2802,"hidden":931},{"src":2800},"https://i0.wp.com/techinformed.com/wp-content/uploads/2024/11/Firefly-a-man-organising-data-its-lit-up-and-he-is-using-his-fingers-in-the-air-in-an-office-the-1.jpg?fit=2688%2C1536&ssl=1",[1112],"2025-02-27","/news/becoming-ai-savvy-for-transformation",{"title":2778,"description":2797},{"loc":2803},"9.news/846.becoming-ai-savvy-for-transformation","gWDcy0MGEksx-k_cWfdJOMrk0ZHHAJvBCcMkfqfrz4U",{"id":2809,"title":2810,"author":2811,"body":2816,"description":201,"extension":923,"meta":2830,"navigation":931,"path":2835,"seo":2836,"sitemap":2837,"stem":2838,"__hash__":2839},"content/9.news/847.pathway-mentioned-in-the-financial-times.md","Forbes: Pathway Navigates Next Road For AI Foundational Models",{"name":2812,"description":2813,"img":2814,"provider":10,"website":2815},"Forbes","Adrian Bridgwater - Senior Contributor","/assets/blog/avatars/forbes-av.png","https://www.forbes.com/",{"type":12,"value":2817,"toc":2828},[2818,2820,2826],[2612,2819,2615],{"id":2614},[20,2821,2618,2822,2623],{},[717,2823,2824],{"href":2824,"rel":2825},"https://www.forbes.com/sites/adrianbridgwater/2025/02/13/pathway-navigates-next-road-for-ai-foundational-models/",[733],[2625,2827],{"url":2824},{"title":201,"searchDepth":510,"depth":510,"links":2829},[],{"redirection":931,"thumbnail":2831,"tags":2833,"date":2834},{"src":2832},"https://imageio.forbes.com/specials-images/imageserve/67ac8673cfa548308522a6f4/Park-System-In-Pennsylvania-Town/960x0.jpg?format=jpg&width=1440",[1112],"2025-02-13","/news/pathway-mentioned-in-the-financial-times",{"title":2810,"description":201},{"loc":2835},"9.news/847.pathway-mentioned-in-the-financial-times","O17X0F02Q9ww1Jsw8ymcbpiX4hRncW2Bn7YidYvKhJE",{"id":2841,"title":2842,"author":2843,"body":2850,"description":201,"extension":923,"meta":2864,"navigation":931,"path":2869,"seo":2870,"sitemap":2871,"stem":2872,"__hash__":2873},"content/9.news/848.pathway-ceo-predicts-2025-ai-trends.md","Pathway CEO and co-founder predicts 2025 AI trends: Will your startup survive the shift?",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},"zuzanna","zuzanna-stamirowska","Zuzanna Stamirowska","CEO","/assets/authors/zuzanna-stamirowska.png","https://www.linkedin.com/in/stamirowska/",{"type":12,"value":2851,"toc":2862},[2852,2854,2860],[2612,2853,2615],{"id":2614},[20,2855,2618,2856,2623],{},[717,2857,2858],{"href":2858,"rel":2859},"https://techfundingnews.com/pathway-ceo-and-co-founder-predicts-2025-ai-trends-will-your-startup-survive-the-shift/",[733],[2625,2861],{"url":2858},{"title":201,"searchDepth":510,"depth":510,"links":2863},[],{"layout":925,"redirection":931,"thumbnail":2865,"tags":2867,"date":2868},{"src":2866},"https://techfundingnews.com/wp-content/uploads/2024/11/pathway.jpg",[1112],"2024-12-19","/news/pathway-ceo-predicts-2025-ai-trends",{"title":2842,"description":201},{"loc":2869},"9.news/848.pathway-ceo-predicts-2025-ai-trends","a_lg9ZKqG4a3pL0gWgxXDnlacNe-76lzdwULUYI6EjA",{"id":2875,"title":2876,"author":2877,"body":2881,"description":201,"extension":923,"meta":2895,"navigation":931,"path":2899,"seo":2900,"sitemap":2901,"stem":2902,"__hash__":2903},"content/9.news/849.pathway-featured-maddyness-insights-and-predictions.md","Pathway featured in Maddyness 2025 Insights and Predictions",{"name":2878,"img":2879,"provider":10,"website":2880},"Maddyness","/assets/blog/avatars/maddyness-avatar.png","https://www.maddyness.com/",{"type":12,"value":2882,"toc":2893},[2883,2885,2891],[2612,2884,2615],{"id":2614},[20,2886,2618,2887,2623],{},[717,2888,2889],{"href":2889,"rel":2890},"https://www.maddyness.com/uk/2024/12/20/prompts-and-predictions-part-2-startup-founders-share-their-insights-and-ambitions-for-2025/",[733],[2625,2892],{"url":2889},{"title":201,"searchDepth":510,"depth":510,"links":2894},[],{"layout":925,"redirection":931,"thumbnail":2896,"tags":2898,"date":1195,"hidden":931},{"src":2897,"provider":10},"/assets/blog/thumbnails/maddyness-prediction-th.png",[1112],"/news/pathway-featured-maddyness-insights-and-predictions",{"title":2876,"description":201},{"loc":2899},"9.news/849.pathway-featured-maddyness-insights-and-predictions","_y_KmTa8LGsKC3IhPjHj9R6LK69CYYoiWdHcBHcrqyc",{"id":2905,"title":2906,"author":2907,"body":2911,"description":201,"extension":923,"meta":2925,"navigation":931,"path":2930,"seo":2931,"sitemap":2932,"stem":2933,"__hash__":2934},"content/9.news/850.cnbc-india-spotlighting-pathway.md","CNBC India spotlighting Pathway",{"name":2908,"description":201,"website":2909,"img":2910,"provider":10},"cnbctv18","https://www.cnbctv18.com","/assets/blog/avatars/cnbctv18-av.png",{"type":12,"value":2912,"toc":2923},[2913,2915,2921],[2612,2914,2615],{"id":2614},[20,2916,2618,2917,2623],{},[717,2918,2919],{"href":2919,"rel":2920},"https://www.cnbctv18.com/business/startup/ai-startup-pathway-raises-10-million-dollar-seed-funding-19520684.htm",[733],[2625,2922],{"url":2919},{"title":201,"searchDepth":510,"depth":510,"links":2924},[],{"layout":925,"redirection":931,"thumbnail":2926,"tags":2928,"date":2929,"hidden":931},{"src":2927,"provider":10},"/assets/blog/thumbnails/cnbc-india-spotlighting-pathway-th.jpg",[1112],"2024-12-06","/news/cnbc-india-spotlighting-pathway",{"title":2906,"description":201},{"loc":2930},"9.news/850.cnbc-india-spotlighting-pathway","1xjNVkt5kb5PvGv4WLobrtc3_SkGq2SC75JY_4tswMI",{"id":2936,"title":2937,"author":2938,"body":2939,"description":201,"extension":923,"meta":2952,"navigation":931,"path":2957,"seo":2958,"sitemap":2959,"stem":2960,"__hash__":2961},"content/9.news/851.female-founded-pathway-raises-10m-to-power-future-of-live-ai-systems.md","Pathway raises $10 million in seed funding round",{"name":2908,"description":201,"website":2909,"img":2910,"provider":10},{"type":12,"value":2940,"toc":2950},[2941,2943,2948],[2612,2942,2615],{"id":2614},[20,2944,2618,2945,2623],{},[717,2946,2919],{"href":2919,"rel":2947},[733],[2625,2949],{"url":2919},{"title":201,"searchDepth":510,"depth":510,"links":2951},[],{"layout":925,"redirection":931,"thumbnail":2953,"tags":2955,"date":2956,"hidden":931},{"src":2954},"https://images.cnbctv18.com/uploads/2024/06/untitled-design-12-2024-06-e6878307a9dc2dc2aa80d08efe758942.jpg?impolicy=website&width=640&height=360",[1112],"2024-12-02","/news/female-founded-pathway-raises-10m-to-power-future-of-live-ai-systems",{"title":2937,"description":201},{"loc":2957},"9.news/851.female-founded-pathway-raises-10m-to-power-future-of-live-ai-systems","8ku4BkZNWzkv240suHS55ojf0gKHcb_DNW0uxHESOdo",{"id":2963,"title":2964,"author":2965,"body":2966,"description":2980,"extension":923,"meta":2981,"navigation":931,"path":2985,"seo":2986,"sitemap":2987,"stem":2988,"__hash__":2989},"content/9.news/852.pathway-10m-seed-round-news.md","Parisian AI startup Pathway on moving to the US: 'We need to be in the room where it happens, and it happens in the Bay Area",{"name":2606,"description":201,"website":2607,"img":2608,"provider":10},{"type":12,"value":2967,"toc":2978},[2968,2970,2976],[2612,2969,2615],{"id":2614},[20,2971,2618,2972,2623],{},[717,2973,2974],{"href":2974,"rel":2975},"https://sifted.eu/articles/pathway-10m-seed-round-news",[733],[2625,2977],{"url":2974},{"title":201,"searchDepth":510,"depth":510,"links":2979},[],"The Paris-founded startup has just closed a $10m seed round and is planning to double down on its US presence",{"layout":925,"redirection":931,"thumbnail":2982,"tags":2984,"date":2956,"hidden":931},{"src":2983,"provider":10},"/assets/blog/thumbnails/pathway-10m-seed-round-news-th.png",[1112],"/news/pathway-10m-seed-round-news",{"title":2964,"description":2980},{"loc":2985},"9.news/852.pathway-10m-seed-round-news","JtD4UiK73sDn8tYLvb7StqfmZBJ9eOzRGp21uSGE-hY",{"id":2991,"title":2992,"author":2993,"body":2996,"description":3010,"extension":923,"meta":3011,"navigation":931,"path":3016,"seo":3017,"sitemap":3018,"stem":3019,"__hash__":3020},"content/9.news/852.pathway-raises-10-million-in-funding-to-advance-the-development-of-live-ai.md","ETCIO Southeast Asia covers Pathway Seed Round",{"name":2994,"description":201,"img":2995,"provider":10},"ET CIOSEA","/assets/blog/avatars/et-ciosea-av.png",{"type":12,"value":2997,"toc":3008},[2998,3000,3006],[2612,2999,2615],{"id":2614},[20,3001,2618,3002,2623],{},[717,3003,3004],{"href":3004,"rel":3005},"https://ciosea.economictimes.indiatimes.com/amp/news/corporate/pathway-raises-10-million-in-funding-to-advance-the-development-of-live-ai/115953873",[733],[2625,3007],{"url":3004},{"title":201,"searchDepth":510,"depth":510,"links":3009},[],"The funding will enable the next step in Pathway’s mission to create LiveAI™ systems capable of complex reasoning",{"layout":925,"redirection":931,"thumbnail":3012,"tags":3014,"date":3015,"hidden":931},{"src":3013,"provider":10},"/assets/blog/thumbnails/et-cio-th.png",[1112],"2024-12-04","/news/pathway-raises-10-million-in-funding-to-advance-the-development-of-live-ai",{"title":2992,"description":3010},{"loc":3016},"9.news/852.pathway-raises-10-million-in-funding-to-advance-the-development-of-live-ai","QWhtZvXoC83HxBpz394uPKJDrAs4pkqaV9Nyt1RbgQk",{"id":3022,"title":3023,"author":3024,"body":3028,"description":3042,"extension":923,"meta":3043,"navigation":931,"path":3048,"seo":3049,"sitemap":3050,"stem":3051,"__hash__":3052},"content/9.news/853.LLM-series-Pathway-Taking-LLMs-out-of-pilot-into-production.md","LLM series - Pathway: Taking LLMs out of pilot into production",{"name":3025,"description":201,"website":3026,"img":3027,"provider":10},"ComputerWeekly","https://www.computerweekly.com","/assets/blog/avatars/computer-weekly-av.png",{"type":12,"value":3029,"toc":3040},[3030,3032,3038],[2612,3031,2615],{"id":2614},[20,3033,2618,3034,2623],{},[717,3035,3036],{"href":3036,"rel":3037},"https://www.computerweekly.com/blog/CW-Developer-Network/LLM-series-Pathway-Taking-LLMs-out-of-pilot-into-production",[733],[2625,3039],{"url":3036},{"title":201,"searchDepth":510,"depth":510,"links":3041},[],"Pathway bills itself as the ultimate data processing framework for the AI era",{"layout":925,"redirection":931,"thumbnail":3044,"tags":3046,"date":3047,"hidden":931},{"src":3045,"provider":10},"/assets/blog/thumbnails/computerweekly-th.png",[1112],"2024-01-08","/news/llm-series-pathway-taking-llms-out-of-pilot-into-production",{"title":3023,"description":3042},{"loc":3048},"9.news/853.LLM-series-Pathway-Taking-LLMs-out-of-pilot-into-production","e6OqnXu9gI_X0qFCAJZKgHG5ac8Xx9IJM8ZfA-kBLMQ",{"id":3054,"title":3055,"author":3056,"body":3060,"description":3074,"extension":923,"meta":3075,"navigation":931,"path":3080,"seo":3081,"sitemap":3082,"stem":3083,"__hash__":3084},"content/9.news/854.investors-ceos-founders-chatgpt-journey.md","Industry Leaders Comment On Biggest Lessons From ChatGPT’s Journey So Far",{"name":3057,"description":201,"website":3058,"img":3059,"provider":10},"TechRound","https://www.techround.com","/assets/blog/avatars/techround-av.png",{"type":12,"value":3061,"toc":3072},[3062,3064,3070],[2612,3063,2615],{"id":2614},[20,3065,2618,3066,2623],{},[717,3067,3068],{"href":3068,"rel":3069},"https://techround.co.uk/news/investors-ceos-founders-chatgpt-journey/",[733],[2625,3071],{"url":3068},{"title":201,"searchDepth":510,"depth":510,"links":3073},[],"CEOs, founders and investors have stepped in to share their reflections, findings and learnings from the AI startup’s journey since its launch in 2022. Running an AI startup isn’t easy, but it can be successful, as we’ve seen with OpenAI. Here’s what experts think",{"layout":925,"redirection":931,"thumbnail":3076,"tags":3078,"date":3079,"hidden":931},{"src":3077},"https://techround.co.uk/wp-content/uploads/fly-images/120424/pramod-tiwari-QPWKc779h2E-unsplash-scaled-e1732897235475-1600x1159.jpg",[1112],"2024-11-29","/news/investors-ceos-founders-chatgpt-journey",{"title":3055,"description":3074},{"loc":3080},"9.news/854.investors-ceos-founders-chatgpt-journey","ksx6oXnZroL4TNNG-euP3qkRKxvMLHT018-znUW3k0c",{"id":3086,"title":3087,"author":3088,"body":3092,"description":3106,"extension":923,"meta":3107,"navigation":931,"path":3111,"seo":3112,"sitemap":3113,"stem":3114,"__hash__":3115},"content/9.news/855.as-cohere-and-writer-mine-the-live-ai-arena-pathway-joins-the-pack-with-a-10m-round.md","As Cohere and Writer mine the ‘LiveAI™’ arena, Pathway joins the pack with a $10M round",{"name":3089,"description":201,"website":3090,"img":3091,"provider":10},"TechCrunch","https://www.techcrunch.com","/assets/blog/avatars/techcrunch-av.png",{"type":12,"value":3093,"toc":3104},[3094,3096,3102],[2612,3095,2615],{"id":2614},[20,3097,2618,3098,2623],{},[717,3099,3100],{"href":3100,"rel":3101},"https://techcrunch.com/2024/11/29/as-cohere-and-writer-mine-the-live-ai-arena-pathway-joins-the-pack-with-a-10m-round/",[733],[2625,3103],{"url":3100},{"title":201,"searchDepth":510,"depth":510,"links":3105},[],"Pathway named as one of the French fastest-growing companies in 2023",{"layout":925,"redirection":931,"thumbnail":3108,"tags":3110,"date":3079},{"src":3109,"provider":10},"/assets/blog/thumbnails/techcrunch-art-th.png",[1112],"/news/as-cohere-and-writer-mine-the-live-ai-arena-pathway-joins-the-pack-with-a-10m-round",{"title":3087,"description":3106},{"loc":3111},"9.news/855.as-cohere-and-writer-mine-the-live-ai-arena-pathway-joins-the-pack-with-a-10m-round","mpIRgm4CNdcJUjHVwqTwySYijUxF5XMOMehbbryKtj0",{"id":3117,"title":3118,"author":3119,"body":3122,"description":3160,"extension":923,"meta":3161,"navigation":931,"path":3165,"seo":3166,"sitemap":3167,"stem":3168,"__hash__":3169},"content/9.news/856.jsec-pathway-ai-collaboration-steadfast-foxtrot-2024.md","Joint Support and Enabling Command collaborates with AI company Pathway to combine industry and military expertise",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},"pathway-team","Pathway Team",{"type":12,"value":3123,"toc":3158},[3124,3127,3134,3140,3143,3146,3149,3152,3155],[2612,3125,3118],{"id":3126},"joint-support-and-enabling-command-collaborates-with-ai-company-pathway-to-combine-industry-and-military-expertise",[20,3128,3129],{},[3130,3131],"img",{"alt":3132,"src":3133},"JSEC, NATO, Allied Command Transformation, Pathway logos","/assets/content/blog/jsec-pathway-banner.png",[20,3135,3136,3139],{},[36,3137,3138],{},"Ulm, Germany. 01 OCTOBER 2024"," From 11 to 18 September, more than 250 participants from 24 nations and various NATO entities engaged in one of the largest military enablement exercises at the Joint Support and Enabling Command (JSEC).",[20,3141,3142],{},"Steadfast Foxtrot 2024 not only trained experts in enablement, reinforcement by forces and sustainment but also set the stage for unveiling NATO’s steps towards the next generation of data processing and simulation systems in close collaboration with the Artificial Intelligence (AI) company Pathway.",[20,3144,3145],{},"“Robust and innovative data processing technology such as delivered by Pathway, unlocks new capabilities for critical use cases at scale,” emphasizes Major General Gerry Ewart-Brookes, Deputy Chief of Staff Plans. The ability to combine military data sources and open-source information such as civil traffic, social media alerts, and media is crucial for the planning and execution of military operations.",[20,3147,3148],{},"With its functional demonstrator, the Reinforcement Enablement Simulation Tool (REST), Pathway developed the cornerstone for further development of AI-supported solutions to NATO.",[20,3150,3151],{},"According to Major General Dirk Kipper, Deputy Chief of Staff Operations, the smart combination of NATO and open source data will speed up situational awareness and bring it to the necessary level, required to successfully operate in the 21st century.",[20,3153,3154],{},"Exercise Steadfast Foxtrot 2024 tested NATO’s resilience in the face of a greater menace at the Eastern European borders. Military personnel from Allied nations together with NATO staff trained to strengthen mutual cooperation and test the sustainability of NATO forces. Anticipating the movement of troops and equipment from the east coast of North America across the Atlantic and the European continent has never been so critical for an effective deterrence of Allied territory.",[20,3156,3157],{},"This initiative is a great example of how NATO aims to bridge the gap between what the industry can offer, and the military expertise in order to further improve the safety of the Alliance’s one billion citizens.",{"title":201,"searchDepth":510,"depth":510,"links":3159},[],"Discover how Joint Support and Enabling Command (JSEC) collaborates with AI company Pathway during Steadfast Foxtrot 2024 to advance NATO's data processing and simulation capabilities, enhancing military operations and resilience in Eastern Europe",{"layout":925,"thumbnail":3162,"tags":3164,"date":1234},{"src":3163,"provider":10},"/assets/blog/thumbnails/jsec-pathway-th.png",[1112,1021],"/news/jsec-pathway-ai-collaboration-steadfast-foxtrot-2024",{"title":3118,"description":3160},{"loc":3165},"9.news/856.jsec-pathway-ai-collaboration-steadfast-foxtrot-2024","cVEuEEcGva-IWzGMlhvEqvWwJqa1RpGqT51yM5oC_do",{"id":3171,"title":3172,"author":3173,"body":3174,"description":3205,"extension":923,"meta":3206,"navigation":931,"path":3211,"seo":3212,"sitemap":3213,"stem":3214,"__hash__":3215},"content/9.news/857.pathway-available-on-aws-cloud.md","Pathway is available on AWS Cloud!",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":3175,"toc":3203},[3176,3179,3188,3196],[20,3177,3178],{},"We are excited to share that Pathway is now available on AWS Marketplace!",[20,3180,3181,3182,3187],{},"Pathway BYOL (Bring Your Own License) ",[717,3183,3186],{"href":3184,"rel":3185},"https://aws.amazon.com/marketplace/pp/prodview-qijbgoyohele4",[733],"Container"," is now available on AWS Marketplace. The listing offers a ready-to-use Docker image with Pathway and all its dependencies pre-installed within the AWS ecosystem.",[20,3189,3190,3191,3195],{},"You can start using the container now. A free license key will help you unlock more ",[717,3192,3194],{"href":3193},"/pricing","features"," of the Pathway framework.",[20,3197,3198,3199,3202],{},"Don’t know where to start? Try our ",[717,3200,1066],{"href":3201},"/developers/user-guide/deployment/aws-fargate-deploy#running-pathway-program-in-aws-cloud-with-fargate"," on building an ETL process that tracks GitHub commit history, removes sensitive data, and loads the results into a Delta Lake.",{"title":201,"searchDepth":510,"depth":510,"links":3204},[],"Discover Pathway on AWS Marketplace! Access the Pathway BYOL container with pre-installed dependencies, ready for deployment in the AWS ecosystem. Unlock more features with a free license key and explore our tutorial on building an ETL process using Pathway",{"layout":925,"thumbnail":3207,"tags":3209,"date":3210,"hidden":931},{"src":3208,"provider":10},"/assets/blog/thumbnails/pathway-is-available-on-aws-cloud-th.png",[1112],"2024-08-16","/news/pathway-available-on-aws-cloud",{"title":3172,"description":3205},{"loc":3211},"9.news/857.pathway-available-on-aws-cloud","B8cxL_Elr1700f5Q7XwcKQwqfKphzudoZLq1rVZtLS4",{"id":3217,"title":3218,"author":3219,"body":3220,"description":3290,"extension":923,"meta":3291,"navigation":931,"path":3296,"seo":3297,"sitemap":3298,"stem":3299,"__hash__":3300},"content/9.news/858.pathway-meetup-2024.md","The Future of Large Language Models by Lukasz Kaiser and Jan Chorowski",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":3221,"toc":3287},[3222,3225,3228,3237,3241,3246,3249,3254,3257,3268,3271,3274],[2612,3223,3218],{"id":3224},"the-future-of-large-language-models-by-lukasz-kaiser-and-jan-chorowski",[20,3226,3227],{},"In April 2024, Pathway hosted an incredible meetup in San Francisco, bringing together some of the brightest minds in AI and data science.",[20,3229,3230,3231,3236],{},"We welcomed Łukasz Kaiser, co-author of ",[717,3232,3235],{"href":3233,"rel":3234},"https://arxiv.org/abs/1706.03762",[733],"\"Attention is All You Need\""," and Jan Chorowski, Pathway’s CTO, who shared their vision on the future of Large Language Models and the roadmap towards more intelligent foundational Large Language Models (LLMs). Joined by many senior developers, architects, and founders working on generative AI projects, they discussed the evolution of deep learning, the role of Reinforcement Learning with Human Feedback, the future of LLMs, and how achieving infinite LLM Context Windows can be made possible through innovative engineering and efficient retrieval mechanisms.",[15,3238,3240],{"id":3239},"key-topics-covered-by-lukasz-kaiser-and-jan-chorowski","Key Topics Covered by Lukasz Kaiser and Jan Chorowski:",[268,3242,3243],{},[72,3244,3245],{},"Role of Retrievers in Reinforcement Learning for Intelligent LLMs\nŁukasz Kaiser, a renowned researcher at OpenAI who is a co-author of TensorFlow and Transformer Architecture as well as core contributor of Open AI’s GPT-4 and ChatGPT, explored the evolution and future of deep learning technologies and their future.",[20,3247,3248],{},"He emphasized that more data and compute lead to better results but highlighted the impending data scarcity. Łukasz discussed how in the future, training with fewer, high-quality retrieved data points will be the key to enhancing LLM performance. He also explained the importance of powerful retrieval mechanisms, integrating personal and organizational knowledge graphs, and efficient context provisioning for effective Reinforcement Learning with Human Feedback. Additionally, Łukasz mentioned a missed observation on parsing from his seminal paper \"Attention is All You Need,\" and shared his vision for future Large Language Models (LLMs).",[268,3250,3251],{"start":897},[72,3252,3253],{},"How Retrievers and LLMs Help Each Other and Achieving Infinite LLM Context Windows",[20,3255,3256],{},"Jan Chorowski, CTO of Pathway and a prominent figure in AI and NLP, extended the discussion by focusing on the essential role of context and retrieval in AI systems.",[20,3258,3259,3260,3263,3264,3267],{},"Building on Łukasz Kaiser's insights, he highlighted the \"yin and yang” relationship between Large Language Models (LLMs) and retrieval systems. Effective LLM performance and reinforcement learning require robust retrieval mechanisms, and efficient retrieval relies on the processing power of LLMs. Jan shared an ",[717,3261,3262],{"href":1245},"example of Adaptive Retrieval Augmented Generation (RAG)"," where they achieved great accuracy at a quarter of the cost by leveraging LLM preprocessing. He emphasized the need for tighter integration to achieve infinite LLM Context Windows and cost-effective AI solutions, inviting the audience to explore these concepts further with resources and examples at ",[717,3265,3266],{"href":719},"Pathway's developer site",".",[20,3269,3270],{},"Watch the recording:",[171,3272],{"src":3273},"https://www.youtube.com/watch?v=_7VirEqCZ4g",[69,3275,3276,3279],{},[72,3277,3278],{},"​Talk 1: Deep Learning Past and Future: What Comes After GPT? by Lukasz Kaiser (“Attention is All you Need” co-author, Senior Researcher at OpenAI)",[72,3280,3281,3282,3286],{},"Talk 2: Taming Unstructured Data: Which Indexing Strategy Wins? by Jan Chorowski (CTO, ",[717,3283,3285],{"href":3284},"/","Pathway",")",{"title":201,"searchDepth":510,"depth":510,"links":3288},[3289],{"id":3239,"depth":510,"text":3240},"Key to Reinforcement Learning (RL) in LLMs as data gets scarce? Insights from Transformer co-inventor and Pathway's CTO at our Bay Area Meetup",{"layout":925,"thumbnail":3292,"tags":3294,"date":3295},{"src":3293,"provider":10},"/assets/blog/thumbnails/pathway-meetup-th.jpg",[1112],"2024-04-30","/news/pathway-meetup-2024",{"title":3218,"description":3290},{"loc":3296},"9.news/858.pathway-meetup-2024","FiRw-86KNah1fnIOBFP78zbKyAHFfP2Nc-yxW5TP6PQ",{"id":3302,"title":3106,"author":3303,"body":3307,"description":3106,"extension":923,"meta":3321,"navigation":931,"path":3326,"seo":3327,"sitemap":3328,"stem":3329,"__hash__":3330},"content/9.news/859.fw-500-startups-2024.md",{"name":3304,"description":201,"website":3305,"img":3306,"provider":10},"FrenchWeb","https://www.frenchweb.fr/","/assets/blog/avatars/frenchweb-av.png",{"type":12,"value":3308,"toc":3319},[3309,3311,3317],[2612,3310,2615],{"id":2614},[20,3312,2618,3313,2623],{},[717,3314,3315],{"href":3315,"rel":3316},"https://dataroom.frenchweb.fr/frenchweb500-startup-2024",[733],[2625,3318],{"url":3315},{"title":201,"searchDepth":510,"depth":510,"links":3320},[],{"layout":925,"redirection":931,"thumbnail":3322,"tags":3324,"date":3325,"hidden":931},{"src":3323,"provider":10},"/assets/blog/thumbnails/fw-500-startups-th.png",[1112],"2024-03-01","/news/fw-500-startups-2024",{"title":3106,"description":3106},{"loc":3326},"9.news/859.fw-500-startups-2024","WXLC7wCDVUr72lKUxMjQEK5bLN7MwD81G5jDyi-Rf90",{"id":3332,"title":3333,"author":3334,"body":3338,"description":3352,"extension":923,"meta":3353,"navigation":931,"path":3358,"seo":3359,"sitemap":3360,"stem":3361,"__hash__":3362},"content/9.news/860.tldr-pathway.md","Pathway highlighted in the TLDR newsletter",{"name":3335,"description":201,"website":3336,"img":3337,"provider":10},"TLDR.tech","https://tldr.tech/","/assets/blog/avatars/tldr-av.png",{"type":12,"value":3339,"toc":3350},[3340,3342,3348],[2612,3341,2615],{"id":2614},[20,3343,2618,3344,2623],{},[717,3345,3346],{"href":3346,"rel":3347},"https://tldr.tech/tech/2023-12-22",[733],[2625,3349],{"url":3346},{"title":201,"searchDepth":510,"depth":510,"links":3351},[],"Pathway’s GitHub Repo has been shared in the “Quick Links” section by the TDLR newsletter",{"layout":925,"redirection":931,"thumbnail":3354,"tags":3356,"date":3357,"hidden":931},{"src":3355,"provider":10},"/assets/blog/thumbnails/tldr-th.png",[1112],"2023-12-22","/news/tldr-pathway",{"title":3333,"description":3352},{"loc":3358},"9.news/860.tldr-pathway","kGSmToAAjt6ftzBn3VUCmMoM9q4-ujyzaLFS1DlKbJI",{"id":3364,"title":3365,"author":3366,"body":3370,"description":3384,"extension":923,"meta":3385,"navigation":931,"path":3390,"seo":3391,"sitemap":3392,"stem":3393,"__hash__":3394},"content/9.news/861.unified-real-time-platforms.md","Pathway positioned in the Unified Real-Time Platforms category",{"name":3367,"description":201,"linkedin":3368,"img":3369,"provider":10},"RTInsights","https://www.rtinsights.com/author/three-amigos/","/assets/blog/avatars/rtinsights-av.png",{"type":12,"value":3371,"toc":3382},[3372,3374,3380],[2612,3373,2615],{"id":2614},[20,3375,2618,3376,2623],{},[717,3377,3378],{"href":3378,"rel":3379},"https://www.rtinsights.com/unified-real-time-platforms/",[733],[2625,3381],{"url":3378},{"title":201,"searchDepth":510,"depth":510,"links":3383},[],"Unified Real-Time Platforms (URPs) are a new category of software designed to handle demanding applications that deal with both streaming data and data at rest",{"layout":925,"redirection":931,"thumbnail":3386,"tags":3388,"date":3389,"hidden":931},{"src":3387,"provider":10},"/assets/blog/thumbnails/unified-realtime-platforms-th.png",[1112],"2024-01-21","/news/unified-real-time-platforms",{"title":3365,"description":3384},{"loc":3390},"9.news/861.unified-real-time-platforms","tUeXjBznawmae6S-QTQalwOh0HDa8xhFhQpSLQ7Jr1A",{"id":3396,"title":3397,"author":3398,"body":3399,"description":3413,"extension":923,"meta":3414,"navigation":931,"path":3419,"seo":3420,"sitemap":3421,"stem":3422,"__hash__":3423},"content/9.news/862.cdo-magazine.md","How Businesses Can Create Data Frameworks for Real-world AI",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":3400,"toc":3411},[3401,3403,3409],[2612,3402,2615],{"id":2614},[20,3404,2618,3405,2623],{},[717,3406,3407],{"href":3407,"rel":3408},"https://www.cdomagazine.tech/aiml/how-businesses-can-create-data-frameworks-for-real-world-ai?utm_content=277910962&utm_medium=social&utm_source=linkedin&hss_channel=lcp-40830869",[733],[2625,3410],{"url":3407},{"title":201,"searchDepth":510,"depth":510,"links":3412},[],"How to make #RealTimeAI a reality? Zuzanna Stamirowska sheds light on the potential of unifying batch, streaming, and data workflows to derive value from data in real time for timely decisions",{"redirection":931,"thumbnail":3415,"tags":3417,"date":3418,"hidden":931},{"src":3416,"provider":10},"/assets/blog/thumbnails/cdo-magazine-th.png",[1112],"2024-02-09","/news/cdo-magazine",{"title":3397,"description":3413},{"loc":3419},"9.news/862.cdo-magazine","46Hhlp7iLHc2mwsuyzRcLpFM0HhI9pfOOUIjviV_vb0",{"id":3425,"title":3426,"author":3427,"body":3430,"description":3444,"extension":923,"meta":3445,"navigation":931,"path":3450,"seo":3451,"sitemap":3452,"stem":3453,"__hash__":3454},"content/9.news/863.blef-podcast.md","Pathway mentioned by 'Blef' in the DataGen Podcast around top data trends for 2024",{"name":3428,"description":201,"img":3429,"provider":10},"Data Gen Podcast","/assets/blog/avatars/datagen-av.jpg",{"type":12,"value":3431,"toc":3442},[3432,3434,3440],[2612,3433,2615],{"id":2614},[20,3435,2618,3436,2623],{},[717,3437,3438],{"href":3438,"rel":3439},"https://www.youtube.com/watch?v=vEguK-J2QIg",[733],[2625,3441],{"url":3438},{"title":201,"searchDepth":510,"depth":510,"links":3443},[],"Christophe Blefari (aka Blef.fr) is a Data Engineer and the author of the best-known data newsletter in the ecosystem: Blef.fr",{"redirection":931,"thumbnail":3446,"tags":3448,"date":3449,"lang":1580,"hidden":931},{"src":3447,"provider":10},"/assets/blog/thumbnails/blef-podcast-th.jpg",[1112,1239],"2023-12-20","/news/blef-podcast",{"title":3426,"description":3444},{"loc":3450},"9.news/863.blef-podcast","76KIU-SPPxJrjOPvQs_kITaUMicOhiw1xAvmiAHuDlg",{"id":3456,"title":3457,"author":3458,"body":3459,"description":3473,"extension":923,"meta":3474,"navigation":931,"path":3479,"seo":3480,"sitemap":3481,"stem":3482,"__hash__":3483},"content/9.news/863.genai-unleashed.md","Gen AI Unleashed: Trends and Challenges for Enterprises in 2024, by Zuzanna Stamirowska, CEO of Pathway",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":3460,"toc":3471},[3461,3463,3469],[2612,3462,2615],{"id":2614},[20,3464,2618,3465,2623],{},[717,3466,3467],{"href":3467,"rel":3468},"https://itsupplychain.com/gen-ai-unleashed-trends-and-challenges-for-enterprises-in-2024/",[733],[2625,3470],{"url":3467},{"title":201,"searchDepth":510,"depth":510,"links":3472},[],"Whether your enterprise operates in finance, healthcare, manufacturing, or any other sector, the transformative potential of GenAI is simply too significant to be ignored. Discover Zuzanna Stamirowska’s thoughts on Gen AI for 2024.",{"layout":925,"redirection":931,"thumbnail":3475,"tags":3477,"date":3478,"hidden":931},{"src":3476,"provider":10},"/assets/blog/thumbnails/genai-unleashed-th.png",[1112],"2023-12-26","/news/genai-unleashed",{"title":3457,"description":3473},{"loc":3479},"9.news/863.genai-unleashed","J0DMOcqAo8b0MeL8nbNPt0HaNGjq2Dy6JPISxsoXNDM",{"id":3485,"title":3486,"author":3487,"body":3492,"description":3533,"extension":923,"meta":3534,"navigation":931,"path":3539,"seo":3540,"sitemap":3541,"stem":3542,"__hash__":3543},"content/9.news/864.modern-data-stack.md","Client Testimonial: La Poste at Modern Data Stack",{"name":3488,"img":3489,"provider":10,"linkedin":3490,"website":3491},"Modern Data Stack","/assets/blog/avatars/modern-data-stack-av.png","https://www.linkedin.com/company/modern-data-stack-france/","https://www.meetup.com/fr-FR/modern-data-stack-france/",{"type":12,"value":3493,"toc":3531},[3494,3497,3500,3507,3514,3517,3520],[2612,3495,3486],{"id":3496},"client-testimonial-la-poste-at-modern-data-stack",[20,3498,3499],{},"Jean-Paul Fabre, Head of Technological Innovation at the Group La Poste, will present how several analytical use cases - network optimization, asset utilization improvement, flow management, Paris 2024 Olympic Games preparation, etc - are enabled by a digital twin and a data model that combines batch and streaming data thanks to Pathway unified engine.",[20,3501,3502,3506],{},[717,3503,3488],{"href":3504,"rel":3505},"https://www.linkedin.com/company/modern-data-stack-france/posts/?feedView=all",[733]," is a community for knowledge-sharing and networking around data thanks to cutting-edge tech.",[20,3508,3509,3510,3513],{},"In January, ",[717,3511,3488],{"href":3504,"rel":3512},[733]," will host in Criteo’s offices in Paris around streaming and the modern data stack.",[20,3515,3516],{},"In addition to La Poste, Decathlon, Michelin, BPCE, OVH Cloud and Christophe Blefari will also share their experience and use cases.",[20,3518,3519],{},"Sign up for the event on January 31st, in Paris - it’s free. Let us know if you are coming!",[20,3521,3522],{},[717,3523,3530],{"href":3524,"rel":3525,"className":3526},"https://docs.google.com/forms/d/e/1FAIpQLSd7R-EUtGDZtvknd5ImTrSE754XhY96KlZeY5Qd_8A9tfekkA/viewform",[733],[3527,3528,3529],"button","button--secondary","button--secondary-text","Sign up for the event",{"title":201,"searchDepth":510,"depth":510,"links":3532},[],"Optimizing Colissimo flows in real time with Pathway. Testimonial",{"layout":925,"thumbnail":3535,"tags":3537,"date":3538,"hidden":931},{"src":3536,"provider":10},"/assets/blog/thumbnails/modern-data-stack-news-th.png",[1112],"2023-12-01","/news/modern-data-stack",{"title":3486,"description":3533},{"loc":3539},"9.news/864.modern-data-stack","JVrpzBm_S2Sqxz_UyEuCLa-auivi9kutTmqtAY1Tvio",{"id":3545,"title":3546,"author":3547,"body":3551,"description":3606,"extension":923,"meta":3607,"navigation":931,"path":3616,"seo":3617,"sitemap":3618,"stem":3619,"__hash__":3620},"content/9.news/865.pathway-best-data-orchestration-supply-chain.md","Pathway named as one of the best solutions in Data Orchestration in Supply chain",{"name":3548,"img":3549,"provider":10,"website":3550},"Daphni Venture Capital","/assets/blog/avatars/daphni-av.png","https://www.daphni.com/",{"type":12,"value":3552,"toc":3604},[3553,3556,3559,3562,3565,3568,3594,3597],[2612,3554,3546],{"id":3555},"pathway-named-as-one-of-the-best-solutions-in-data-orchestration-in-supply-chain",[20,3557,3558],{},"While demands for more analytics-driven and real-time decision making remain high, technical hurdles limit user adoption and opportunities to drive business outcomes from these solutions. Pathway has been identified as one of the best solutions in Transportation management systems by Daphni Venture Capital.",[20,3560,3561],{},"Daphni Venture Capital is a Paris-based Venture Capital (VC) firm focusing on supporting early-stage tech startups, known for its commitment to fostering innovation in the digital space. Notably, Daphni has invested in transformative companies such as Aircall and the quantum company PASQAL.",[20,3563,3564],{},"Pathway is the fastest data processing engine supporting unified workflows for batch, streaming data, and LLM applications",[20,3566,3567],{},"At Pathway we are proud to enable real-time intelligence in Logistics and Supply Chain. With Pathway, get value in under 24 hours: gather your data, get immediately a coherent data model you can work with, and access insights on the fly.\nRead more about how Pathway has been designed for",[69,3569,3570,3576,3582,3588],{},[72,3571,3572],{},[717,3573,3575],{"href":3574},"/solutions/logistics#operations","Operations",[72,3577,3578],{},[717,3579,3581],{"href":3580},"/solutions/logistics#iot-deployment-experts","IoT deployment experts",[72,3583,3584],{},[717,3585,3587],{"href":3586},"/solutions/logistics#risk-insurance-security","Risk, Insurance and Security teams",[72,3589,3590],{},[717,3591,3593],{"href":3592},"/solutions/logistics#digital-data-teams","Digital & Data teams",[20,3595,3596],{},"… to address business problems in logistics and supply chain at scale.\nPathway already works with leaders in the market, such as CMA CGM, DB Schenker or La Poste.",[20,3598,3599,3600],{},"Read the full article on Daphni: ",[717,3601,3602],{"href":3602,"rel":3603},"https://www.daphni.com/insights/data-orchestration-in-supply-chain/",[733],{"title":201,"searchDepth":510,"depth":510,"links":3605},[],"According to Gartner analysts Christian Titze and Noha Tohamy, 'By 2026, 50% of organizations will have to evaluate analytics and business intelligence (ABI) and data science and machine learning (DSML) platforms as a single platform due to market convergence'",{"layout":925,"thumbnail":3608,"tags":3610,"date":3611,"related":3612,"hidden":931},{"src":3609,"provider":10},"/assets/blog/thumbnails/daphni-th.jpg",[1112],"2023-11-27",[3613,3614,3615],"/blog/gartner","/news/gartner-market-guide-supply-chain","/blog/market-guide-event-stream-processing","/news/pathway-best-data-orchestration-supply-chain",{"title":3546,"description":3606},{"loc":3616},"9.news/865.pathway-best-data-orchestration-supply-chain","4ghOSAtXlqWHNJEeiEw7yUELIij_IiOo6Ldx6dwV0J0",{"id":3622,"title":3623,"author":3624,"body":3628,"description":3642,"extension":923,"meta":3643,"navigation":931,"path":3647,"seo":3648,"sitemap":3649,"stem":3650,"__hash__":3651},"content/9.news/866.eu-startup-news.md","Pathway named among the Top Startups Transforming the European business landscape",{"name":3625,"description":201,"img":3626,"provider":10,"website":3627},"EU Startup News","/assets/blog/avatars/eustartup-news-avatar.png","https://eustartup.news/",{"type":12,"value":3629,"toc":3640},[3630,3632,3638],[2612,3631,2615],{"id":2614},[20,3633,2618,3634,2623],{},[717,3635,3636],{"href":3636,"rel":3637},"https://eustartup.news/which-french-b2b-startups-are-transforming-the-european-business-landscape/",[733],[2625,3639],{"url":3636},{"title":201,"searchDepth":510,"depth":510,"links":3641},[],"Which French B2B Startups Are Transforming the European Business Landscape?",{"layout":925,"redirection":931,"thumbnail":3644,"tags":3646,"date":1496,"hidden":931},{"src":3645,"provider":10},"/assets/blog/thumbnails/eustartup-news-th.png",[1112],"/news/eu-startup-news",{"title":3623,"description":3642},{"loc":3647},"9.news/866.eu-startup-news","rYlBzCuz5KQDJI3Yq7omWvailj9OkmWVBOlyP50Yibw",{"id":3653,"title":3654,"author":3655,"body":3659,"description":3714,"extension":923,"meta":3715,"navigation":931,"path":3720,"seo":3721,"sitemap":3722,"stem":3723,"__hash__":3724},"content/9.news/867.gartner-a-n-di-solutions.md","Pathway is featured as a best-suited vendor candidate for Analytics and Decision Intelligence solutions for Supply Chain by Gartner",{"name":3656,"img":3657,"website":3658},"Gartner","/assets/content/blog/gartner-avatar.png","https://www.gartner.com/myhomepage",{"type":12,"value":3660,"toc":3712},[3661,3665,3668,3676,3679,3697,3700,3703],[2612,3662,3664],{"id":3663},"pathway-was-selected-as-a-vendor-candidate-for-analytics-decision-intelligence-adi-for-supply-chain","Pathway was selected as a vendor candidate for analytics & decision intelligence (A&DI) for Supply Chain.",[20,3666,3667],{},"This Tool released by Gartner has been designed to support supply chain technology leaders in identifying suitable and best-fit supply chain analytics and decision intelligence vendor candidates for their software evaluation process.",[20,3669,3670,3671,3675],{},"At Pathway we are proud to ",[717,3672,3674],{"href":3673},"/solutions/logistics","enable real-time intelligence in Logistics"," and Supply Chain. With Pathway, get value in under 24 hours: gather your data, get immediately a coherent data model you can work with, and access insights on the fly.",[20,3677,3678],{},"Read more about how Pathway has been designed for",[69,3680,3681,3685,3689,3693],{},[72,3682,3683],{},[717,3684,3575],{"href":3574},[72,3686,3687],{},[717,3688,3581],{"href":3580},[72,3690,3691],{},[717,3692,3587],{"href":3586},[72,3694,3695],{},[717,3696,3593],{"href":3592},[20,3698,3699],{},"… to address business problems in logistics and supply chain at scale.",[20,3701,3702],{},"Pathway already works with leaders in the market, such as CMA CGM, DB Schenker or La Poste.",[20,3704,3705,3706,3711],{},"For Gartner clients, feel free to download the full Excel spreadsheet ",[717,3707,3710],{"href":3708,"rel":3709},"https://bit.ly/3SqkBj0",[733],"Tool: Identify A&DI Solutions for Supply Chain",", and reach out!",{"title":201,"searchDepth":510,"depth":510,"links":3713},[],"Gartner just released its Tool to support supply chain technology leaders in identifying suitable and best-fit supply chain analytics and decision intelligence (A&DI) vendors candidates",{"layout":925,"thumbnail":3716,"tags":3717,"date":3718,"enterprise":931,"related":3719,"hidden":931},{"src":1742,"provider":10},[1112],"2023-10-23",[3613,3614,3615],"/news/gartner-a-n-di-solutions",{"title":3654,"description":3714},{"loc":3720},"9.news/867.gartner-a-n-di-solutions","W86AT8CM-3jbPwhxicvJ4YOEpyJxxH2f3j6wTqTJo1M",{"id":3726,"title":3727,"author":3728,"body":3731,"description":3745,"extension":923,"meta":3746,"navigation":931,"path":3751,"seo":3752,"sitemap":3753,"stem":3754,"__hash__":3755},"content/9.news/868.tech-informed-article.md","A coffee with… Zuzanna Stamirowska",{"name":2780,"description":201,"img":3729,"provider":10,"website":3730},"/assets/blog/avatars/techinformed-avatar.png","https://techinformed.com/",{"type":12,"value":3732,"toc":3743},[3733,3735,3741],[2612,3734,2615],{"id":2614},[20,3736,2618,3737,2623],{},[717,3738,3739],{"href":3739,"rel":3740},"https://techinformed.com/a-coffee-with-zuzanna-stamirowska/",[733],[2625,3742],{"url":3739},{"title":201,"searchDepth":510,"depth":510,"links":3744},[],"The strategy polymath and chief exec of real-time data analytics firm Pathway on data pipeline complexity, injecting privacy into large language models and meeting Nobel Prize-winning mathematician John Nash",{"layout":925,"redirection":931,"thumbnail":3747,"tags":3749,"date":3750,"hidden":931},{"src":3748,"provider":10},"/assets/blog/thumbnails/coffe-with-th.png",[1112],"2023-10-11","/news/tech-informed-article",{"title":3727,"description":3745},{"loc":3751},"9.news/868.tech-informed-article","0szxkJa6F5YgejX5x7aM8sM0yQFOC4Ry6XspxXN21Ww",{"id":3757,"title":3758,"author":3759,"body":3760,"description":3774,"extension":923,"meta":3775,"navigation":931,"path":3780,"seo":3781,"sitemap":3782,"stem":3783,"__hash__":3784},"content/9.news/869.european-financial-review.md","Building Data Frameworks for Real-time AI Applications",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":3761,"toc":3772},[3762,3764,3770],[2612,3763,2615],{"id":2614},[20,3765,2618,3766,2623],{},[717,3767,3768],{"href":3768,"rel":3769},"https://www.europeanfinancialreview.com/building-data-frameworks-for-real-time-ai-applications/",[733],[2625,3771],{"url":3768},{"title":201,"searchDepth":510,"depth":510,"links":3773},[],"Zuzanna Stamirowska, Co-founder and CEO of Pathway, wrote an article explaining the new paradigm of real-time AI applications for financial services organizations, which hold the promise of delivering faster, smart and more efficient processes.",{"redirection":931,"thumbnail":3776,"tags":3778,"date":3779,"hidden":931},{"src":3777,"provider":10},"/assets/blog/thumbnails/european-financial-review-th.png",[1112],"2023-09-24","/news/european-financial-review",{"title":3758,"description":3774},{"loc":3780},"9.news/869.european-financial-review","fCHz0-hkdrUrwwiMr7H6BRNOh-qbOaVufmBzqVArkQ4",{"id":3786,"title":3787,"author":3788,"body":3792,"description":3806,"extension":923,"meta":3807,"navigation":931,"path":3812,"seo":3813,"sitemap":3814,"stem":3815,"__hash__":3816},"content/9.news/871.wearewomen-article.md","Enabling AI to unlearn and self-correct like a human",{"name":3789,"description":201,"img":3790,"website":3791},"We Are Tech Women","/assets/content/blog/avatars/wearetechwomen-avatar.png","https://wearetechwomen.com/",{"type":12,"value":3793,"toc":3804},[3794,3796,3802],[2612,3795,2615],{"id":2614},[20,3797,2618,3798,2623],{},[717,3799,3800],{"href":3800,"rel":3801},"https://wearetechwomen.com/enabling-ai-to-unlearn-and-self-correct-iike-a-human/",[733],[2625,3803],{"url":3800},{"title":201,"searchDepth":510,"depth":510,"links":3805},[],"Pathway developed the breakthrough capability to combine batch and streaming logic in the same workflow, AI systems can now be continuously trained or updated with new streaming data, with revisions made to certain data points without requiring a full batch data upload",{"layout":925,"redirection":931,"thumbnail":3808,"tags":3810,"date":3811,"hidden":931},{"src":3809,"provider":10},"/assets/blog/thumbnails/wearetechwomen-th.png",[1112],"2023-08-31","/news/wearewomen-article",{"title":3787,"description":3806},{"loc":3812},"9.news/871.wearewomen-article","n-jTcCcD9tgvQmXMbF6uRL9oHXW3Phg-qUi1Hroi5jU",{"id":3818,"title":3819,"author":3820,"body":3821,"description":3835,"extension":923,"meta":3836,"navigation":931,"path":3841,"seo":3842,"sitemap":3843,"stem":3844,"__hash__":3845},"content/9.news/872.iot-for-all-handle-out-of-order-data-in-your-iot-pipeline.md","How to Handle Out-of-Order Data in Your IoT Pipeline",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":3822,"toc":3833},[3823,3825,3831],[2612,3824,2615],{"id":2614},[20,3826,2618,3827,2623],{},[717,3828,3829],{"href":3829,"rel":3830},"https://www.iotforall.com/how-to-handle-out-of-order-data-in-your-iot-pipeline",[733],[2625,3832],{"url":3829},{"title":201,"searchDepth":510,"depth":510,"links":3834},[],"If some of your #data looks odd, it's possible that something went wrong in your #IoT data pipeline. Pathway explores out-of-order data, one of the most vexing IoT data issues in today's streaming systems, including: The causes, Fixing data errors and Out-of-order logging",{"layout":925,"redirection":931,"thumbnail":3837,"tags":3839,"date":3840,"hidden":931},{"src":3838,"provider":10},"/assets/blog/thumbnails/iotforall-th.png",[1112],"2023-08-21","/news/iot-for-all-handle-out-of-order-data-in-your-iot-pipeline",{"title":3819,"description":3835},{"loc":3841},"9.news/872.iot-for-all-handle-out-of-order-data-in-your-iot-pipeline","Vkmh87Qmtb7853QAq8hv48zMrU7EvX9cIgT12XnWj9A",{"id":3847,"title":3848,"author":3849,"body":3854,"description":3868,"extension":923,"meta":3869,"navigation":931,"path":3874,"seo":3875,"sitemap":3876,"stem":3877,"__hash__":3878},"content/9.news/873.les-echos-deeptech.md","Pathway quoted in Les Echos: Deeptech - the answer to tomorrow's challenges",{"name":3850,"description":3851,"img":3852,"website":3853},"Les Echos","Economic and financial news from France","/assets/content/blog/LesEchos_icon.webp","https://www.lesechos.fr/",{"type":12,"value":3855,"toc":3866},[3856,3858,3864],[2612,3857,2615],{"id":2614},[20,3859,2618,3860,2623],{},[717,3861,3862],{"href":3862,"rel":3863},"https://www.lesechos.fr/idees-debats/cercle/opinion-la-deeptech-est-la-reponse-aux-defis-de-demain-1972505",[733],[2625,3865],{"url":3862},{"title":201,"searchDepth":510,"depth":510,"links":3867},[],"Despite the uncertainty surrounding its financial viability, French VC partner, Pierre-Eric Leibovici, believes that deeptech can enable entrepreneurs to play a key role in resolving environmental and societal challenges",{"layout":925,"redirection":931,"lang":1580,"thumbnail":3870,"tags":3872,"date":3873,"hidden":931},{"src":3871,"provider":10},"/assets/blog/thumbnails/lesechos-th.png",[1112],"2023-08-25","/news/les-echos-deeptech",{"title":3848,"description":3868},{"loc":3874},"9.news/873.les-echos-deeptech","itJBiHUJHpr4yDPzfdCjO0aTvJP5s5wTORCzYu-Bk_8",{"id":3880,"title":3881,"author":3882,"body":3887,"description":3901,"extension":923,"meta":3902,"navigation":931,"path":3907,"seo":3908,"sitemap":3909,"stem":3910,"__hash__":3911},"content/9.news/874.financial-times-skeptical-case.md","Pathway quoted in the FT: The skeptical case on generative AI",{"name":3883,"description":3884,"img":3885,"website":3886},"Financial Times","Worldʼs leading global business publication","/assets/content/blog/financial-times-avatar.png","https://www.ft.com/",{"type":12,"value":3888,"toc":3899},[3889,3891,3897],[2612,3890,2615],{"id":2614},[20,3892,2618,3893,2623],{},[717,3894,3895],{"href":3895,"rel":3896},"https://www.ft.com/content/ed323f48-fe86-4d22-8151-eed15581c337",[733],[2625,3898],{"url":3895},{"title":201,"searchDepth":510,"depth":510,"links":3900},[],"John Thornhill highlights how investors are betting on companies that can deploy #GenAI models to solve real-world problems, giving the French startup Pathway as an example",{"redirection":931,"layout":925,"thumbnail":3903,"tags":3905,"date":3906,"hidden":931},{"src":3904,"provider":10},"/assets/blog/thumbnails/financial-times-th.png",[1112],"2023-08-17","/news/financial-times-skeptical-case",{"title":3881,"description":3901},{"loc":3907},"9.news/874.financial-times-skeptical-case","QIaIsM2AiPeXBIcdOdQenHsMmnqSdqE7wNPh0YVZVxY",{"id":3913,"title":3914,"author":3915,"body":3920,"description":3934,"extension":923,"meta":3935,"navigation":931,"path":3940,"seo":3941,"sitemap":3942,"stem":3943,"__hash__":3944},"content/9.news/875.techeu-realtime-value-to-logistics.md","Pathway bring real-time value to logistics through machine unlearning",{"name":3916,"description":3917,"img":3918,"website":3919},"tech.eu","European Startup, Funding and Technology News","/assets/content/blog/techeu-avatar.png","https://tech.eu/",{"type":12,"value":3921,"toc":3932},[3922,3924,3930],[2612,3923,2615],{"id":2614},[20,3925,2618,3926,2623],{},[717,3927,3928],{"href":3928,"rel":3929},"https://tech.eu/2023/08/22/pathway-machine-unlearning/",[733],[2625,3931],{"url":3928},{"title":201,"searchDepth":510,"depth":510,"links":3933},[],"Pathway innovates real-time data processing, combating outdated or inaccurate data in AI models",{"redirection":931,"layout":925,"thumbnail":3936,"tags":3938,"date":3939,"hidden":931},{"src":3937,"provider":10},"/assets/blog/thumbnails/techeu-th.png",[1112],"2023-08-22","/news/techeu-realtime-value-to-logistics",{"title":3914,"description":3934},{"loc":3940},"9.news/875.techeu-realtime-value-to-logistics","WsBuBxZA6mUkj0GqJfOWebk1ycLQMERD0Gb8JB5g-mo",{"id":3946,"title":3947,"author":3948,"body":3949,"description":3963,"extension":923,"meta":3964,"navigation":931,"path":3969,"seo":3970,"sitemap":3971,"stem":3972,"__hash__":3973},"content/9.news/876.la-revue-ia-article.md","LLM and real-time learning article for 'La revue IA'",{"id":2644,"url":2645,"name":2646,"description":2647,"img":2648,"provider":10,"linkedin":2649},{"type":12,"value":3950,"toc":3961},[3951,3953,3959],[2612,3952,2615],{"id":2614},[20,3954,2618,3955,2623],{},[717,3956,3957],{"href":3957,"rel":3958},"https://larevueia.fr/llm-et-apprentissage-en-temps-reel/",[733],[2625,3960],{"url":3957},{"title":201,"searchDepth":510,"depth":510,"links":3962},[],"Claire Nouet, COO at Pathway wrote an article about LLM and real-time learning describing limits of static models, vector and index database utility for LLM models, and Pathway LLM App",{"redirection":931,"layout":925,"lang":1580,"thumbnail":3965,"tags":3967,"date":3968,"hidden":931},{"src":3966,"provider":10},"/assets/blog/thumbnails/la-revue-th.png",[1112],"2023-08-27","/news/la-revue-ia-article",{"title":3947,"description":3963},{"loc":3969},"9.news/876.la-revue-ia-article","y3Gnbde2Zk3NSchYpKJp4TSKGWjyT6SL2Z8lnU-6UPg",{"id":3975,"title":3976,"author":3977,"body":3979,"description":4034,"extension":923,"meta":4035,"navigation":931,"path":4040,"seo":4041,"sitemap":4042,"stem":4043,"__hash__":4044},"content/9.news/877.maddyness-gen-ai-mapping.md","Pathway named as a promising Generative AI leader (in French)",{"name":2878,"img":3978,"website":2880},"/assets/content/blog/maddyness-avatar.png",{"type":12,"value":3980,"toc":4032},[3981,3984,3997,4005,4012,4015,4021,4025],[2612,3982,3976],{"id":3983},"pathway-named-as-a-promising-generative-ai-leader-in-french",[20,3985,3986,3987,3992,3993,3286],{},"The generative AI market was worth almost $40 billion in 2022 and should approach $70 billion by the end of this year, according to ",[717,3988,3991],{"href":3989,"rel":3990},"https://www.bloomberg.com/company/press/generative-ai-to-become-a-1-3-trillion-market-by-2032-research-finds/",[733],"Bloomberg Intelligence",". And this is just the beginning, as the market is expected to reach $1,300 billion by 2032 (Source: ",[717,3994,3996],{"href":3989,"rel":3995},[733],"Bloomberg: Generative AI to Become a $1.3 Trillion Market by 2032, Research Finds.",[20,3998,3999,4004],{},[717,4000,4003],{"href":4001,"rel":4002},"https://www.resonance.vc/",[733],"Resonance Venture",", a French Venture Capital, released a mapping of the main GenAI players in France, including established French companies such as Hugging Face.",[20,4006,4007,4011],{},[717,4008,3285],{"href":4009,"rel":4010},"https://pathway.com/",[733]," is the single, integrated processing layer for real-time intelligence. It allows easy mix-and-match of batch, streaming, and LLM architectures - all within one engine.",[20,4013,4014],{},"Real-time learning is made possible by an effective and scalable engine, which powers LLMs and machine learning models. These models are automatically updated thanks to a framework that combines streaming and batch data, and which is user-friendly and flexible for developers, data engineers, and data scientists. Leading experts in the field of artificial intelligence make up the team, which is headed by Zuzanna Stamirowska. They include CTO Jan Chorowski, co-authors of Geoff Hinton and Yoshua Bengio, as well as Business Angel Lukasz Kaiser, who co-authored Tensor Flow and is also known as the \"T\" in ChatGPT.",[20,4016,4017,4020],{},[36,4018,4019],{},"Zuzanna Stamirowska, CEO & Co-Founder of Pathway",", comments: “Our mission has been to enable real-time data processing, while giving developers a simple experience regardless of whether they work with batch, streaming, or LLM systems. Pathway is truly facilitating the convergence of historical and real-time data for the first time.”",[198,4022],{":zoomable":200,"alt":4023,"src":4024},"A list of companies in the ecosystem where Pathway has a place in data preparation","assets/content/blog/ecosysteme-gen-ai-francais.png",[20,4026,4027,4028],{},"Read the full article on Maddyness:\n",[717,4029,4030],{"href":4030,"rel":4031},"https://www.maddyness.com/2023/07/21/france-europe-ia-generative/",[733],{"title":201,"searchDepth":510,"depth":510,"links":4033},[],"Maddyness reposted a mapping of future European GenAI leaders",{"layout":925,"thumbnail":4036,"tags":4038,"date":4039,"hidden":931},{"src":4037},"/assets/content/blog/maddyness-gen-ai-mapping-th.png",[1112],"2023-07-26","/news/maddyness-gen-ai-mapping",{"title":3976,"description":4034},{"loc":4040},"9.news/877.maddyness-gen-ai-mapping","Eihm6ie265YMlNhHk1f07Qd20bUSgCAPSi1bGAZB4Gw",{"id":4046,"title":4047,"author":4048,"body":4049,"description":4063,"extension":923,"meta":4064,"navigation":931,"path":4068,"seo":4069,"sitemap":4070,"stem":4071,"__hash__":4072},"content/9.news/878.maddyness-article-about-pathway.md","French deep tech start-up announces the general launch of its data processing engine",{"name":2878,"img":2879,"provider":10,"website":2880},{"type":12,"value":4050,"toc":4061},[4051,4053,4059],[2612,4052,2615],{"id":2614},[20,4054,2618,4055,2623],{},[717,4056,4057],{"href":4057,"rel":4058},"https://www.maddyness.com/2023/07/26/pathway-ia/",[733],[2625,4060],{"url":4057},{"title":201,"searchDepth":510,"depth":510,"links":4062},[],"Comment Pathway veut permettre aux IA d’apprendre et «d’oublier» en temps réel",{"redirection":931,"lang":1580,"thumbnail":4065,"tags":4067,"date":4039,"hidden":931},{"src":4066},"/assets/content/blog/maddyness-th.png",[1112],"/news/maddyness-article-about-pathway",{"title":4047,"description":4063},{"loc":4068},"9.news/878.maddyness-article-about-pathway","6bw1TzK3-BpHJ_7S92d75rmNoDj1q17ICFs0s1cz7Yc",{"id":4074,"title":4075,"author":4076,"body":4080,"description":4094,"extension":923,"meta":4095,"navigation":931,"path":4099,"seo":4100,"sitemap":4101,"stem":4102,"__hash__":4103},"content/9.news/879.nextweb-article.md","AI startup launches ‘fastest data processing engine’ on the market",{"name":4077,"img":4078,"linkedin":4079},"The Next Web","/assets/content/blog/thenextweb-avatar.png","https://thenextweb.com/",{"type":12,"value":4081,"toc":4092},[4082,4084,4090],[2612,4083,2615],{"id":2614},[20,4085,2618,4086,2623],{},[717,4087,4088],{"href":4088,"rel":4089},"https://thenextweb.com/news/ai-startup-launches-fastest-data-processing-engine-market",[733],[2625,4091],{"url":4088},{"title":201,"searchDepth":510,"depth":510,"links":4093},[],"Female-led Pathway says its system can 'forget' in real-time, like a human",{"redirection":931,"thumbnail":4096,"tags":4098,"date":4039,"hidden":931},{"src":4097},"/assets/content/blog/thenextweb-th.png",[1112],"/news/nextweb-article",{"title":4075,"description":4094},{"loc":4099},"9.news/879.nextweb-article","C8KvjMP7iX7Mb6iHM6TrsS7HoN3kLvL_NofgN_RZmfY",{"id":4105,"title":4106,"author":4107,"body":4108,"description":4137,"extension":923,"meta":4138,"navigation":931,"path":4143,"seo":4144,"sitemap":4145,"stem":4146,"__hash__":4147},"content/9.news/880.200-startups-disrupting-supply-chains.md","Pathway Named as a Top Startup Disrupting Supply Chains",{"id":2644,"url":2645,"name":2646,"description":2647,"img":2648,"provider":10,"linkedin":2649},{"type":12,"value":4109,"toc":4135},[4110,4114,4117,4120],[2612,4111,4113],{"id":4112},"pathway-is-listed-as-a-top-start-up-disrupting-supply-chains","Pathway is listed as a top start-up disrupting supply chains.",[20,4115,4116],{},"Pathway has been named in the Management and optimization category, notably thanks to its public success stories with leading logistics companies and supply chain departments.",[20,4118,4119],{},"Pathway Logistics App is the lighthouse data platform built in the Pathway framework. The one-stop-shop cloud-based application is providing immediately actionable insights on top of data for logistics assets, including IoT data and status data.",[20,4121,4122,4123,4127,4128,4130,4131,3267],{},"Discover our success stories with ",[717,4124,4126],{"href":4125},"/success-stories/db-schenker","DB Schenker",", ",[717,4129,1085],{"href":2671}," and ",[717,4132,4134],{"href":4133},"/success-stories/cma-cgm","CMA CGM",{"title":201,"searchDepth":510,"depth":510,"links":4136},[],"Stratégies Logistique listed the 200 top startups disrupting supply chains, and rewarded Pathway in the Management and Optimization category",{"layout":925,"thumbnail":4139,"tags":4141,"date":4142,"hidden":931},{"src":4140},"/assets/content/blog/strategies-logistique-th.png",[1112],"2023-05-11","/news/200-startups-disrupting-supply-chains",{"title":4106,"description":4137},{"loc":4143},"9.news/880.200-startups-disrupting-supply-chains","8QHr8UZiBnsMUC8YH-TnPkRlCqvgnaS0iUcDJmS3Il8",{"id":4149,"title":4150,"author":4151,"body":4154,"description":4168,"extension":923,"meta":4169,"navigation":931,"path":4174,"seo":4175,"sitemap":4176,"stem":4177,"__hash__":4178},"content/9.news/881.le-point.md","Pathway CEO featured in the ranking of the next generation of geniuses by the French national weekly Le Point",{"name":4152,"img":4153},"Le Point","/assets/content/blog/le-point-avatar.png",{"type":12,"value":4155,"toc":4166},[4156,4158,4164],[2612,4157,2615],{"id":2614},[20,4159,2618,4160,2623],{},[717,4161,4162],{"href":4162,"rel":4163},"https://www.lepoint.fr/sciences-nature/palmares-des-inventeurs-du-point-la-releve-du-genie-francais-22-06-2023-2525696_1924.php",[733],[2625,4165],{"url":4162},{"title":201,"searchDepth":510,"depth":510,"links":4167},[],"An exceptional jury (including Alain Aspect, the 2022 Nobel Prize in Physics) has selected Pathway among the teams whose breakthroughs will change our lives.",{"layout":925,"redirection":931,"thumbnail":4170,"tags":4172,"date":4173,"lang":1580,"hidden":931},{"src":4171},"/assets/content/blog/le-point-th.png",[1112],"2023-06-22","/news/le-point",{"title":4150,"description":4168},{"loc":4174},"9.news/881.le-point","OUZIcSjAo36ly-2a3b9vVD5VujQtcdg1FYAt1ZO5Cuo",{"id":4180,"title":4181,"author":4182,"body":4184,"description":4227,"extension":923,"meta":4228,"navigation":931,"path":3614,"seo":4232,"sitemap":4233,"stem":4234,"__hash__":4235},"content/9.news/882.gartner-market-guide-supply-chain.md","Pathway is a Representative Vendor in Gartner 2023 Market Guide for Analytics and Decision Intelligence Platforms in Supply Chain",{"name":3656,"img":3657,"website":4183},"https://www.gartner.com/account/signin?method=initialize&TARGET=https%3A%2F%2Fwww.gartner.com%2Fmyhomepage",{"type":12,"value":4185,"toc":4225},[4186,4190,4193,4196,4209,4216],[2612,4187,4189],{"id":4188},"pathway-was-selected-as-a-representative-vendor-in-the-2023-gartner-market-guide-for-analytics-and-decision-intelligence-platforms-in-supply-chain","Pathway was selected as a Representative Vendor in the 2023 Gartner Market Guide for Analytics and Decision Intelligence Platforms in Supply Chain.",[20,4191,4192],{},"According to Gartner analysts Christian Titze and Noha Tohamy, ”By 2026, 50% of organizations will have to evaluate analytics and business intelligence (ABI) and data science and machine learning (DSML) platforms as a single platform due to market convergence.”",[20,4194,4195],{},"At Pathway we are proud to enable industry leaders to “achieve contextualized, connected, and continuous insights” through:",[69,4197,4198,4203],{},[72,4199,4200,4202],{},[36,4201,3285],{},": the most powerful data processing framework, currently used for real-time anomaly detection, predictive analytics, IoT and logs data observability, recommender systems, and alerting, and which works particularly well with data in motion: data tables, live events data, etc.",[72,4204,4205,4208],{},[36,4206,4207],{},"Pathway Logistics App",": our lighthouse data platform built in the Pathway framework. It is a one-stop-shop cloud-based application to provide immediately actionable insights on top of data for logistics assets, including IoT data and status data.",[20,4210,4211,4212,4215],{},"With “functional teams ",[392,4213,4214],{},"..."," looking to speed up cross-functional decision making on the basis of more near-real-time and broader datasets”, Pathway is best positioned to deliver value to Enterprise clients",[20,4217,4218,4219,4224],{},"For Gartner clients, feel free to read the full ",[717,4220,4223],{"href":4221,"rel":4222},"https://www.gartner.com/document/4478399?ref=solrAll&refval=374406409&",[733],"Gartner Market Guide"," for Analytics and Decision Intelligence Platforms in Supply Chain, and do reach out!",{"title":201,"searchDepth":510,"depth":510,"links":4226},[],"Gartner published its latest edition of its Market Guide for Analytics and Decision Intelligence Platforms in Supply Chain, and named Pathway a Representative Vendor",{"layout":925,"thumbnail":4229,"tags":4230,"date":4231,"enterprise":931,"hidden":931},{"src":1742,"provider":10},[1112],"2023-06-26",{"title":4181,"description":4227},{"loc":3614},"9.news/882.gartner-market-guide-supply-chain","iY5aaP1BB-LlLWHL9cIHys1WV38iLrGhHf_DmxJ_GmI",{"id":4237,"title":4238,"author":4239,"body":4240,"description":4273,"extension":923,"meta":4274,"navigation":931,"path":4278,"seo":4279,"sitemap":4280,"stem":4281,"__hash__":4282},"content/9.news/883.vivatech-by-the-french-prime.md","Pathway awarded at VivaTech by the French Prime Minister Elisabeth Borne",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":4241,"toc":4271},[4242,4245,4248,4251,4255,4262,4265,4268],[2612,4243,4238],{"id":4244},"pathway-awarded-at-vivatech-by-the-french-prime-minister-elisabeth-borne",[20,4246,4247],{},"Pathway is proud to announce that Zuzanna Stamirowska, CEO at Pathway was awarded at Viva Technology, Europe’s biggest tech event, held in Paris, France.",[20,4249,4250],{},"Elisabeth Borne, the French Prime Minister awarded Zuzanna Stamirowska for her performance on stage and the achievements of Pathway as the most powerful data processing framework to power real-time data products and pipelines. This happened a few weeks after the CIO of Goldman Sachs declared that “going from batch to real-time (processing) was like going from printed newspapers to the Internet.\"",[4252,4253],"tweet",{"tweet-url":4254},"https://twitter.com/Elisabeth_Borne/status/1669798321550925837",[20,4256,4257,4258,4261],{},"Pathway was brought to life by a stellar team: the CTO Jan Chorowski worked with the Godfathers of AI, Geoff Hinton, and Yoshua Bengio, the CSO Adrian Kosowski had his Ph.D. at 20 and is a world-class expert in high-scale distributed computing, and ",[717,4259,2846],{"href":2849,"rel":4260},[733]," is the author of the state of the art model for forecasting of maritime trade. Pathway is supported by business angels such as Lukasz Kaiser, known to be behind the “T” in GPT.",[20,4263,4264],{},"“Very soon real-time will become the norm for data processing and it’s a game changer for everybody starting from financial services, Formula 1,  supply chains, online marketing, retail, energy…  the list goes on.” declared Zuzanna Stamirowska during her pitch in front of the Viva Tech assembly.",[20,4266,4267],{},"Watch Pathway Winning Pitch",[171,4269],{"src":4270},"https://www.youtube.com/watch?v=iSRUMsM15uw",{"title":201,"searchDepth":510,"depth":510,"links":4272},[],"The real-time revolution starts now. Zuzanna Stamirowska, CEO at Pathway was awarded by the French Prime Minister at Viva Technology",{"layout":925,"thumbnail":4275,"tags":4277,"date":1540,"hidden":931},{"src":4276},"/assets/content/blog/vivatech-by-the-french-prime-th.jpg",[1112,1238],"/news/vivatech-by-the-french-prime",{"title":4238,"description":4273},{"loc":4278},"9.news/883.vivatech-by-the-french-prime","rwC0oXE8zilyf-Z1rReNA22z_4cCppRnSJvxXNiODIQ",{"id":4284,"title":4285,"author":4286,"body":4289,"description":4304,"extension":923,"meta":4305,"navigation":931,"path":4310,"seo":4311,"sitemap":4312,"stem":4313,"__hash__":4314},"content/9.news/884.devmio.md","Founder profile: Zuzanna Stamirowska opening up to devmio",{"name":4287,"description":201,"img":4288},"Devmio","/assets/content/blog/devmio-avatar.png",{"type":12,"value":4290,"toc":4302},[4291,4293,4299],[2612,4292,2615],{"id":2614},[20,4294,2618,4295,2623],{},[717,4296,4297],{"href":4297,"rel":4298},"https://devm.io/careers/women-in-tech-stamirowska",[733],[2625,4300],{"url":4301},"https://devm.io/careers/women-in-tech-stamirowska/",{"title":201,"searchDepth":510,"depth":510,"links":4303},[],"The key to overcoming challenges is to see them as temporary obstacles",{"layout":925,"redirection":931,"thumbnail":4306,"tags":4308,"date":4309,"hidden":931},{"src":4307},"/assets/content/blog/devmio-th.png",[1112],"2023-05-17","/news/devmio",{"title":4285,"description":4304},{"loc":4310},"9.news/884.devmio","DL2GpXj68R5PEnfz-lYL_Xd7bMmntL4nzO9gsKegCCg",{"id":4316,"title":4317,"author":4318,"body":4321,"description":4335,"extension":923,"meta":4336,"navigation":931,"path":4340,"seo":4341,"sitemap":4342,"stem":4343,"__hash__":4344},"content/9.news/885.paris-saclay.md","Interview for Paris-Saclay",{"name":4319,"description":201,"img":4320},"Paris-Saclay","/assets/content/blog/paris-saclay-avatar.png",{"type":12,"value":4322,"toc":4333},[4323,4325,4331],[2612,4324,2615],{"id":2614},[20,4326,2618,4327,2623],{},[717,4328,4329],{"href":4329,"rel":4330},"https://epa-paris-saclay.fr/actualites-et-decryptages/toutes-nos-publications/traitement-de-donnees-en-temps-reel-la-voie-pathway/",[733],[2625,4332],{"url":4329},{"title":201,"searchDepth":510,"depth":510,"links":4334},[],"Traitement de données en temps réel, la voie Pathway",{"layout":925,"redirection":931,"thumbnail":4337,"tags":4339,"date":1564,"lang":1580,"hidden":931},{"src":4338},"/assets/content/blog/paris-saclay-th.png",[1112],"/news/paris-saclay",{"title":4317,"description":4335},{"loc":4340},"9.news/885.paris-saclay","O251ObVwyoWAaDWKZxIy2oUkJsFD6lvwL8etWTRTpM4",{"id":4346,"title":4347,"author":4348,"body":4352,"description":4366,"extension":923,"meta":4367,"navigation":931,"path":4372,"seo":4373,"sitemap":4374,"stem":4375,"__hash__":4376},"content/9.news/886.la-jaune-et-la-rouge.md","La Poste shared their IoT roadmap, and how Pathway helps them with their strategic objectives",{"name":4349,"description":4350,"img":4351},"La Jaune et la Rouge","Polytechnique Alumni","/assets/content/blog/la-june-avatar.png",{"type":12,"value":4353,"toc":4364},[4354,4356,4362],[2612,4355,2615],{"id":2614},[20,4357,2618,4358,2623],{},[717,4359,4360],{"href":4360,"rel":4361},"https://www.lajauneetlarouge.com/nous-travaillons-avec-liot-depuis-longtemps/",[733],[2625,4363],{"url":4360},{"title":201,"searchDepth":510,"depth":510,"links":4365},[],"Interview with Jean-Paul Fabre, Technology Manager at the Innovation and Information Systems department of the Mail & Parcel Services division of La Poste",{"layout":925,"redirection":931,"thumbnail":4368,"tags":4370,"date":4371,"lang":1580,"hidden":931},{"src":4369},"/assets/content/blog/la-june-th.png",[1112],"2023-04-01","/news/la-jaune-et-la-rouge",{"title":4347,"description":4366},{"loc":4372},"9.news/886.la-jaune-et-la-rouge","WXhmxkTT4o7oDXcMR5X6fzi3_PLhPjFxkVDNQjSL9fo",{"id":4378,"title":4379,"author":4380,"body":4383,"description":4397,"extension":923,"meta":4398,"navigation":931,"path":4403,"seo":4404,"sitemap":4405,"stem":4406,"__hash__":4407},"content/9.news/887.emilie-magazine-sciences-po.md","Interview for Emile Magazine, Sciences Po",{"name":4381,"img":4382},"Sciences Po","https://upload.wikimedia.org/wikipedia/commons/7/7c/Sciences_Po_Coat_of_arms.png",{"type":12,"value":4384,"toc":4395},[4385,4387,4393],[2612,4386,2615],{"id":2614},[20,4388,2618,4389,2623],{},[717,4390,4391],{"href":4391,"rel":4392},"https://www.emilemagazine.fr/article/2023/3/1/entretien-pathway",[733],[2625,4394],{"url":4391},{"title":201,"searchDepth":510,"depth":510,"links":4396},[],"Pathway, quand Sciences Po rime avec deeptechs",{"layout":925,"redirection":931,"thumbnail":4399,"tags":4401,"date":4402,"lang":1580,"hidden":931},{"src":4400,"provider":10},"/assets/pictures/image_zuzanna_and_claire.jpg",[1112],"2023-03-31","/news/emilie-magazine-sciences-po",{"title":4379,"description":4397},{"loc":4403},"9.news/887.emilie-magazine-sciences-po","-a6ZH_ZOQ5XZvLC0c_OHRTK0Cz4vC3AI7O9Ta-Owh4I",{"id":4409,"title":4410,"author":4411,"body":4412,"description":4446,"extension":923,"meta":4447,"navigation":931,"path":4452,"seo":4453,"sitemap":4454,"stem":4455,"__hash__":4456},"content/9.news/889.wavestone-data-ai-radar.md","Pathway featured in Wavestone’s 2022 Data & AI radar",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":4413,"toc":4444},[4414,4418,4421,4430,4434],[2612,4415,4417],{"id":4416},"pathway-featured-in-the-2022-french-data-ai-for-the-industry-radar-put-together-by-wavestone","Pathway featured in the 2022 French “Data & AI For the Industry” Radar, put together by Wavestone.",[20,4419,4420],{},"The French consulting firm Wavestone, an established leader in digital transformation, led a mapping initiative to list the best AI and Data startups which are used by industry players.",[20,4422,4423,4424,4426,4427,4429],{},"Pathway, which is serving clients such as ",[717,4425,1085],{"href":2671}," or ",[717,4428,4126],{"href":4125}," has been selected as one of the leading 11 startups in this category, highlighting the quality of its value proposition for the industry.",[198,4431],{"alt":4432,"src":4433},"Wavestone radar","/assets/content/blog/wavestone-radar.png",[20,4435,4436,4437,4440],{},"Discover the full report here:",[4438,4439],"br",{},[717,4441,4442],{"href":4442,"rel":4443},"https://www.wavestone.com/fr/insight/radar-2022-des-startups-data-et-ia-pour-industrie/",[733],{"title":201,"searchDepth":510,"depth":510,"links":4445},[],"Pathway has been featured in Wavestone’s 2022 Data & AI radar for its value proposition in Logistics and Supply chains.",{"aside":951,"layout":925,"thumbnail":4448,"tags":4450,"date":4451,"hidden":931},{"src":4449},"/assets/content/blog/wavestone.png",[925,1112],"2022-11-04","/news/wavestone-data-ai-radar",{"title":4410,"description":4446},{"loc":4452},"9.news/889.wavestone-data-ai-radar","C1wqV2lxpgdKNjQTnzbFHgY8mfNuLO0EdMkusEhp8io",{"id":4458,"title":4459,"author":4460,"body":4461,"description":4477,"extension":923,"meta":4478,"navigation":931,"path":4483,"seo":4484,"sitemap":4485,"stem":4486,"__hash__":4487},"content/9.news/890.LesEchosdeeptechportrait.md","Pathway in Les Echos - CEO Portrait",{"name":3850,"img":3852},{"type":12,"value":4462,"toc":4475},[4463,4465,4472],[2612,4464,2615],{"id":2614},[20,4466,2618,4467,4471],{},[717,4468,4469],{"href":4469,"rel":4470},"https://www.lesechos.fr/start-up/portraits/ces-chercheurs-qui-ont-decide-de-fonder-une-start-up-dans-la-deeptech-1895133in",[733]," a moment.",[2625,4473],{"url":4474},"https://www.lesechos.fr/start-up/portraits/ces-chercheurs-qui-ont-decide-de-fonder-une-start-up-dans-la-deeptech-1895133",{"title":201,"searchDepth":510,"depth":510,"links":4476},[],"You will be taken to https://www.lesechos.fr/start-up/portraits/ces-chercheurs-qui-ont-decide-de-fonder-une-start-up-dans-la-deeptech-1895133in a moment.",{"layout":925,"redirection":931,"date":4479,"thumbnail":4480,"lang":1580,"tags":4482,"hidden":931},"2023-01-09",{"src":4481,"contain":931},"/assets/content/blog/Les_echos_(logo).svg.png",[1112],"/news/lesechosdeeptechportrait",{"title":4459,"description":4477},{"loc":4483},"9.news/890.LesEchosdeeptechportrait","pZORk4p2FH6u3MrAcCO5EhQ2ELpdd460uq-AmbQUmG8",{"id":4489,"title":4490,"author":4491,"body":4493,"description":4507,"extension":923,"meta":4508,"navigation":931,"path":4513,"seo":4514,"sitemap":4515,"stem":4516,"__hash__":4517},"content/9.news/900.female-led-deeptech-startup.md","Female-led deeptech startup Pathway announces its $4.5m pre-seed round",{"name":2606,"img":4492},"/assets/content/blog/avatars/sifted-av.png",{"type":12,"value":4494,"toc":4505},[4495,4497,4503],[2612,4496,2615],{"id":2614},[20,4498,2618,4499,2623],{},[717,4500,4501],{"href":4501,"rel":4502},"https://sifted.eu/articles/female-led-deeptech-pathway-ai/",[733],[2625,4504],{"url":4501},{"title":201,"searchDepth":510,"depth":510,"links":4506},[],"The French startup helps companies with real-time analysis of fast-moving data",{"layout":925,"redirection":931,"date":4509,"thumbnail":4510,"tags":4512,"hidden":931},"2022-12-06",{"src":4511},"https://images.sifted.eu/wp-content/uploads/2022/12/05160857/Pathway-Zuzanna-CEO-and-Claire-COO-scaled-e1670263090806.jpg?w=2048&h=1054&q=75&fit=crop&auto=compress,format",[1112],"/news/female-led-deeptech-startup",{"title":4490,"description":4507},{"loc":4513},"9.news/900.female-led-deeptech-startup","bWvQ4kdATzaURG-JnB5xocilq2oWQ19d2CtWpvZOr1g",{"id":4519,"title":4520,"author":4521,"body":4524,"description":4538,"extension":923,"meta":4539,"navigation":931,"path":4544,"seo":4545,"sitemap":4546,"stem":4547,"__hash__":4548},"content/9.news/927.Business-cool-interview.md","Zuzanna’s interview for Business Cool ",{"name":4522,"img":4523},"Business Cool","/assets/content/blog/businesscool.png",{"type":12,"value":4525,"toc":4536},[4526,4528,4534],[2612,4527,2615],{"id":2614},[20,4529,2618,4530,2623],{},[717,4531,4532],{"href":4532,"rel":4533},"https://business-cool.com/entreprendre/levee-fonds/pathway-annonce-une-levee-de-fonds-de-45-millions-de-dollars/",[733],[2625,4535],{"url":4532},{"title":201,"searchDepth":510,"depth":510,"links":4537},[],"Pathway annonce une levée de fonds de 4,5 millions de dollars",{"redirection":931,"layout":925,"date":4540,"thumbnail":4541,"tags":4543,"lang":1580,"hidden":931},"2023-01-12",{"src":4542,"provider":10},"/assets/pictures/image_stamirowska_pathway.jpg",[1112],"/news/business-cool-interview",{"title":4520,"description":4538},{"loc":4544},"9.news/927.Business-cool-interview","qzGPfzH066cp8LyQ9Y48ElO_8s0zNsrezjkcnxxwFwA",{"id":4550,"title":4551,"author":4552,"body":4555,"description":4551,"extension":923,"meta":4569,"navigation":931,"path":4574,"seo":4575,"sitemap":4576,"stem":4577,"__hash__":4578},"content/9.news/944.BFMTV.md","Pathway on BFM Business - the French Business TV channel",{"name":4553,"img":4554},"BFM Business","/assets/content/blog/BFM_icon.jpg",{"type":12,"value":4556,"toc":4567},[4557,4559,4565],[2612,4558,2615],{"id":2614},[20,4560,2618,4561,2623],{},[717,4562,4563],{"href":4563,"rel":4564},"https://www.bfmtv.com/economie/replay-emissions/tech-and-co/paris-saclay-spring-2022-quelles-sont-les-cinq-start-up-primees-19-05_VN-202205190687.html",[733],[2625,4566],{"url":4563},{"title":201,"searchDepth":510,"depth":510,"links":4568},[],{"layout":925,"redirection":931,"date":4570,"thumbnail":4571,"tags":4573,"hidden":931},"2022-05-30",{"src":4572,"contain":931},"/assets/content/blog/BFM-Business-Logo.png",[1112],"/news/bfmtv",{"title":4551,"description":4551},{"loc":4574},"9.news/944.BFMTV","F2xlZrHg-lOWCsFl4IVT2Kn7JOokBI_3oWMjtn7qMmU",{"id":4580,"title":4581,"author":4582,"body":4585,"description":201,"extension":923,"meta":4599,"navigation":931,"path":4605,"seo":4606,"sitemap":4607,"stem":4608,"__hash__":4609},"content/9.news/ai-should-think-like-the-human-brain-dragon-hatchling-bdh-copies-neurons-for-unlimited-context-and-higher-efficiency.md","AI should think like the human brain: Dragon Hatchling (BDH) copies neurons for unlimited context and higher efficiency",{"name":4583,"favicon":4584,"website":4584},"Notebook Check","notebookcheck.com",{"type":12,"value":4586,"toc":4597},[4587,4589,4595],[2612,4588,2615],{"id":2614},[20,4590,2618,4591,2623],{},[717,4592,4593],{"href":4593,"rel":4594},"https://www.notebookcheck.com/KI-soll-denken-wie-das-menschliche-Gehirn-Dragon-Hatchling-BDH-kopiert-Neuronen-fuer-unbegrenzten-Kontext-und-hoehere-Effizienz.1142453.0.html",[733],[2625,4596],{"url":4593},{"title":201,"searchDepth":510,"depth":510,"links":4598},[],{"redirection":931,"lang":4600,"tags":4601,"date":4602,"related":951,"thumbnail":4603},"german",[1112,1830],"2025-10-21",{"src":4604,"provider":10},"/assets/blog/thumbnails/notebook-check-th.png","/news/ai-should-think-like-the-human-brain-dragon-hatchling-bdh-copies-neurons-for-unlimited-context-and-higher-efficiency",{"title":4581,"description":201},{"loc":4605},"9.news/ai-should-think-like-the-human-brain-dragon-hatchling-bdh-copies-neurons-for-unlimited-context-and-higher-efficiency","E78jyY34D64jUzRWCgPEw8BIkMSTAf8H0-b_pQya1pw",{"id":4611,"title":4612,"author":4613,"body":4617,"description":4631,"extension":923,"meta":4632,"navigation":931,"path":4637,"seo":4638,"sitemap":4639,"stem":4640,"__hash__":4641},"content/9.news/an-ai-startup-looks-toward-the-post-transformer-era.md","Pathway Looks Toward the Post-Transformer Era",{"name":4614,"img":4615,"provider":10,"website":4616},"Wall Street Journal","/assets/blog/avatars/wsj-th.png","https://www.wsj.com/",{"type":12,"value":4618,"toc":4629},[4619,4621,4627],[2612,4620,2615],{"id":2614},[20,4622,2618,4623,2623],{},[717,4624,4625],{"href":4625,"rel":4626},"https://www.wsj.com/articles/an-ai-startup-looks-toward-the-post-transformer-era-4e362db8",[733],[2625,4628],{"url":4625},{"title":201,"searchDepth":510,"depth":510,"links":4630},[],"The architecture underlying large language models revolutionized AI. Pathway’s Dragon Hatchling is designed to do more",{"layout":925,"redirection":931,"thumbnail":4633,"tags":4635,"date":4636,"pinned":931},{"src":4634},"https://images.wsj.net/im-92775332/social",[1112,1830],"2025-12-01","/news/an-ai-startup-looks-toward-the-post-transformer-era",{"title":4612,"description":4631},{"loc":4637},"9.news/an-ai-startup-looks-toward-the-post-transformer-era","fglFx7Sg1bAkyqpxIht9N7qebCqX7fJIkCYiSMKWpBM",{"id":4643,"title":4644,"author":4645,"body":4648,"description":4662,"extension":923,"meta":4663,"navigation":931,"path":4668,"seo":4669,"sitemap":4670,"stem":4671,"__hash__":4672},"content/9.news/aws-reinvent-2025-the-new-ai-architecture-that-adapts-and-thinks-just-like-humans.md","AWS re:Invent 2025 -The new AI architecture that adapts and thinks just like humans",{"name":4646,"img":4647,"provider":10},"AWS Events","/assets/blog/avatars/aws-av.png",{"type":12,"value":4649,"toc":4660},[4650,4652,4658],[2612,4651,2615],{"id":2614},[20,4653,2618,4654,2623],{},[717,4655,4656],{"href":4656,"rel":4657},"https://www.youtube.com/watch?v=cnUSW0pLFVk",[733],[2625,4659],{"url":4656},{"title":201,"searchDepth":510,"depth":510,"links":4661},[],"This session introduces Baby Dragon Hatchling [BDH], a post-Transformer architecture designed to solve the core barrier faced by today’s AI: its inability to generalize over time",{"redirection":931,"thumbnail":4664,"tags":4666,"date":4667},{"src":4665},"https://img.youtube.com/vi/cnUSW0pLFVk/maxresdefault.jpg",[1112,1830,1238],"2025-12-04","/news/aws-reinvent-2025-the-new-ai-architecture-that-adapts-and-thinks-just-like-humans",{"title":4644,"description":4662},{"loc":4668},"9.news/aws-reinvent-2025-the-new-ai-architecture-that-adapts-and-thinks-just-like-humans","pj1U1xUgGOOG-sGzZuJbk0sbvx6mrV-lANeRewrBiIk",{"id":4674,"title":4675,"author":4676,"body":4677,"description":2629,"extension":923,"meta":4691,"navigation":931,"path":4696,"seo":4697,"sitemap":4698,"stem":4699,"__hash__":4700},"content/9.news/can-ai-learn-and-evolve-like-a-brain-pathways-bold-research-thinks-so.md","Can AI Learn And Evolve Like A Brain? Pathway’s Bold Research Thinks So",{"name":2812,"img":2814,"provider":10,"website":2815},{"type":12,"value":4678,"toc":4689},[4679,4681,4687],[2612,4680,2615],{"id":2614},[20,4682,2618,4683,2623],{},[717,4684,4685],{"href":4685,"rel":4686},"https://www.forbes.com/sites/victordey/2025/10/08/can-ai-learn-and-evolve-like-a-brain-pathways-bold-research-thinks-so/",[733],[2625,4688],{"url":4685},{"title":201,"searchDepth":510,"depth":510,"links":4690},[],{"layout":925,"redirection":931,"thumbnail":4692,"tags":4694,"date":4695,"pinned":931},{"src":4693},"https://imageio.forbes.com/specials-images/imageserve/68e69cf3c94f1ee9ed00f2d3/0x0.jpg?format=jpg&height=900&width=1600&fit=bounds",[1112,1830],"2025-10-08","/news/can-ai-learn-and-evolve-like-a-brain-pathways-bold-research-thinks-so",{"title":4675,"description":2629},{"loc":4696},"9.news/can-ai-learn-and-evolve-like-a-brain-pathways-bold-research-thinks-so","0uC4pPb_VbWHp4Dq3jWS7iAfsy4dkV9VPIsWk25sFqw",{"id":4702,"title":4703,"author":4704,"body":4705,"description":4738,"extension":923,"meta":4739,"navigation":931,"path":4742,"seo":4743,"sitemap":4744,"stem":4745,"__hash__":4746},"content/9.news/data-sommelier-podcast-episode-1-season-2.md","Data Sommelier Podcast Episode 1 Season 2",{"id":2644,"url":2645,"name":2646,"description":2647,"img":2648,"provider":10,"linkedin":2649},{"type":12,"value":4706,"toc":4736},[4707,4710,4722,4730],[20,4708,4709],{},"Data Sommelier Season 2 is kicking off with an episode featuring Pathway! Listen to it:",[20,4711,4712],{},[717,4713,4717],{"href":4714,"rel":4715,"className":4716},"https://www.linkedin.com/events/ai-withlivedata-poweringyourrag7303036682809868288/theater/",[733],[204],[3130,4718],{"alt":4719,"className":4720,"provider":10,"src":4721},"Data Sommelier Season 2 linkedin video",[203],"/assets/blog/thumbnails/data-sommelier-podcast-th.jpg",[20,4723,4724,4725,3267],{},"Listen to other Data Sommelier podcasts (in French or English) on ",[717,4726,4729],{"href":4727,"rel":4728},"https://datanosco.com/datasommelier/",[733],"Datanosco’s website",[2668,4731,4733],{"href":4732},"/success-stories",[20,4734,4735],{},"Learn more about our Success Stories",{"title":201,"searchDepth":510,"depth":510,"links":4737},[],"Data Sommelier Season 2 is kicking off with an episode featuring Pathway! Listen to it",{"layout":925,"thumbnail":4740,"tags":4741,"date":1051,"hidden":931},{"src":4721,"provider":10},[1112,1239],"/news/data-sommelier-podcast-episode-1-season-2",{"title":4703,"description":4738},{"loc":4742},"9.news/data-sommelier-podcast-episode-1-season-2","4-bQcSggB58b5AGoVUipwtx2rabr_lCBH1AEuQKG1lY",{"id":4748,"title":4749,"author":4750,"body":4751,"description":4765,"extension":923,"meta":4766,"navigation":931,"path":4770,"seo":4771,"sitemap":4772,"stem":4773,"__hash__":4774},"content/9.news/embracing-modern-live-data-pipelines-is-key-to-scaling-enterprise-ai.md","Embracing Modern Live Data Pipelines is Key to Scaling Enterprise AI",{"name":3367,"description":201,"linkedin":3368,"img":3369,"provider":10},{"type":12,"value":4752,"toc":4763},[4753,4755,4761],[2612,4754,2615],{"id":2614},[20,4756,2618,4757,2623],{},[717,4758,4759],{"href":4759,"rel":4760},"https://www.rtinsights.com/embracing-modern-live-data-pipelines-is-key-to-scaling-enterprise-ai/",[733],[2625,4762],{"url":4759},{"title":201,"searchDepth":510,"depth":510,"links":4764},[],"Scaling LLM deployments from pilot to production requires and emphasis on adopting real-time, adaptive data frameworks",{"layout":925,"redirection":931,"thumbnail":4767,"tags":4769,"date":1921,"hidden":931},{"src":4768},"https://www.rtinsights.com/wp-content/uploads/2025/03/Depositphotos_539418084_S-800x534.jpg",[1112],"/news/embracing-modern-live-data-pipelines-is-key-to-scaling-enterprise-ai",{"title":4749,"description":4765},{"loc":4770},"9.news/embracing-modern-live-data-pipelines-is-key-to-scaling-enterprise-ai","OZmaFp48K4mGreSB23tDG2i1thQNT_QQQo9aZaX311M",{"id":4776,"title":4777,"author":4778,"body":4780,"description":201,"extension":923,"meta":4794,"navigation":931,"path":4800,"seo":4801,"sitemap":4802,"stem":4803,"__hash__":4804},"content/9.news/forbes-poland-ceo-profile-in-polish.md","Forbes Poland: CEO profile (in Polish)",{"name":2812,"img":2814,"provider":10,"website":4779},"https://www.forbes.pl/",{"type":12,"value":4781,"toc":4792},[4782,4784,4790],[2612,4783,2615],{"id":2614},[20,4785,2618,4786,2623],{},[717,4787,4788],{"href":4788,"rel":4789},"https://www.forbes.pl/polka-chce-wstrzasnac-dolina-krzemowa-z-jej-produktu-korzystaja-juz-intel-i-nato/7bmkk92",[733],[2625,4791],{"url":4788},{"title":201,"searchDepth":510,"depth":510,"links":4793},[],{"redirection":931,"thumbnail":4795,"tags":4797,"date":4798,"related":951,"lang":4799},{"src":4796},"https://ocdn.eu/pulscms-transforms/1/rVgk9kpTURBXy9iZDgwYjVkZGVmM2E3OGZhMTIxMzdhZTE0MjUzZmQ1MS5qcGeSlQPNAXoAzQXpzQNUkwXNA47NAl_eAAGhMAU",[1112],"2025-03-20","polish","/news/forbes-poland-ceo-profile-in-polish",{"title":4777,"description":201},{"loc":4800},"9.news/forbes-poland-ceo-profile-in-polish","ElXrU-Hsh0hNlficbG1cUyv9NRYA_tB9A8V-WEuMuEM",{"id":4806,"title":4807,"author":4808,"body":4809,"description":201,"extension":923,"meta":4823,"navigation":931,"path":4828,"seo":4829,"sitemap":4830,"stem":4831,"__hash__":4832},"content/9.news/from-data-sure-to-ai-savvy-unlocking-the-next-stage-of-business-transformation.md","From Data-sure To AI-savvy: Unlocking The Next Stage Of Business Transformation",{"id":2644,"url":2645,"name":2646,"description":2647,"img":2648,"provider":10,"linkedin":2649},{"type":12,"value":4810,"toc":4821},[4811,4813,4819],[2612,4812,2615],{"id":2614},[20,4814,2618,4815,2623],{},[717,4816,4817],{"href":4817,"rel":4818},"https://www.techdogs.com/inspire/c-suite-scoops/from-data-sure-to-ai-savvy-unlocking-the-next-stage-of-business-transformation",[733],[2625,4820],{"url":4817},{"title":201,"searchDepth":510,"depth":510,"links":4822},[],{"redirection":931,"thumbnail":4824,"tags":4826,"date":4827,"related":951,"hidden":931},{"src":4825,"provider":10},"/assets/blog/thumbnails/from-data-sure-to-ai-savvy-unlocking-the-next-stage-of-business-transformation-th.png",[1112],"2025-04-17","/news/from-data-sure-to-ai-savvy-unlocking-the-next-stage-of-business-transformation",{"title":4807,"description":201},{"loc":4828},"9.news/from-data-sure-to-ai-savvy-unlocking-the-next-stage-of-business-transformation","Oxm2s0yVWPvn15BMl7KGddOK6mSv6UgexLH-WYn1fyI",{"id":4834,"title":4835,"author":4836,"body":4839,"description":4853,"extension":923,"meta":4854,"navigation":931,"path":4859,"seo":4860,"sitemap":4861,"stem":4862,"__hash__":4863},"content/9.news/how-neolabs-are-betting-against-the-openai-model-and-what-it-means-for-founders.md","How 'Neolabs' Are Betting Against the OpenAI Model and What It Means for Founders",{"name":4837,"img":4838,"provider":10},"Inc.","/assets/blog/avatars/inc-av.png",{"type":12,"value":4840,"toc":4851},[4841,4843,4849],[2612,4842,2615],{"id":2614},[20,4844,2618,4845,2623],{},[717,4846,4847],{"href":4847,"rel":4848},"https://www.inc.com/brett-farmiloe/how-neolabs-are-betting-against-the-openai-model-and-what-it-means-for-founders/91279024",[733],[2625,4850],{"url":4847},{"title":201,"searchDepth":510,"depth":510,"links":4852},[],"AI is moving faster and becoming more diverse than ever. The next competitive advantage may come from a new architecture.",{"redirection":931,"thumbnail":4855,"tags":4857,"date":4858,"related":951},{"src":4856},"https://img-cdn.inc.com/image/upload/f_webp,q_auto,c_fit,w_1024/vip/2025/12/neolabs-ai-models-new-inc.jpg",[1112,1830],"2025-12-21","/news/how-neolabs-are-betting-against-the-openai-model-and-what-it-means-for-founders",{"title":4835,"description":4853},{"loc":4859},"9.news/how-neolabs-are-betting-against-the-openai-model-and-what-it-means-for-founders","xFTp6GdYzTVECOLC2H98vulR2Qk4dBbVF0cO3y_N2V0",{"id":4865,"title":4866,"author":4867,"body":4870,"description":4884,"extension":923,"meta":4885,"navigation":931,"path":4890,"seo":4891,"sitemap":4892,"stem":4893,"__hash__":4894},"content/9.news/inside-pathways-post-transformer-architecture-designed-for-memory-and-on-the-fly-learning.md","Inside Pathway's Post-Transformer Architecture Designed for Memory and On-the-Fly Learning",{"name":4868,"img":4869},"Eye on AI","https://yt3.ggpht.com/ytc/AIdro_mjddd9v-_8K0iqLY0aO7UmCi0yPYKQe-QP48kcTeViIQ=s48-c-k-c0x00ffffff-no-rj",{"type":12,"value":4871,"toc":4882},[4872,4874,4880],[2612,4873,2615],{"id":2614},[20,4875,2618,4876,2623],{},[717,4877,4878],{"href":4878,"rel":4879},"https://www.youtube.com/watch?v=E6WmXnEFDgc",[733],[2625,4881],{"url":4878},{"title":201,"searchDepth":510,"depth":510,"links":4883},[],"This episode dives into why Pathway’s Baby Dragon Hatchling (BDH) might mark the beginning of the post-transformer era in AI",{"redirection":931,"thumbnail":4886,"tags":4888,"date":4889,"related":951},{"src":4887},"https://img.youtube.com/vi/E6WmXnEFDgc/maxresdefault.jpg",[1112,1830],"2026-03-11","/news/inside-pathways-post-transformer-architecture-designed-for-memory-and-on-the-fly-learning",{"title":4866,"description":4884},{"loc":4890},"9.news/inside-pathways-post-transformer-architecture-designed-for-memory-and-on-the-fly-learning","u2jUEoLA6AzenMeIQEr2fmBxonoWrEZ9m7K3cYEmnXE",{"id":4896,"title":4897,"author":4898,"body":4900,"description":201,"extension":923,"meta":4914,"navigation":931,"path":4920,"seo":4921,"sitemap":4922,"stem":4923,"__hash__":4924},"content/9.news/inteligencia-artificial-aprender-cerebro-humano.md","Can an artificial intelligence learn like a human brain does? A startup believes it has achieved this",{"name":4899,"img":2814,"provider":10,"website":2815},"Forbes Argentina",{"type":12,"value":4901,"toc":4912},[4902,4904,4910],[2612,4903,2615],{"id":2614},[20,4905,2618,4906,2623],{},[717,4907,4908],{"href":4908,"rel":4909},"https://es-us.noticias.yahoo.com/inteligencia-artificial-aprender-cerebro-humano-231500859.html",[733],[2625,4911],{"url":4908},{"title":201,"searchDepth":510,"depth":510,"links":4913},[],{"layout":925,"redirection":931,"tags":4915,"date":4916,"lang":4917,"thumbnail":4918},[1112,1830],"2025-09-09","spanish",{"src":4919,"provider":10},"/assets/blog/thumbnails/inteligencia-artificial-aprender-cerebro-humano-th.png","/news/inteligencia-artificial-aprender-cerebro-humano",{"title":4897,"description":201},{"loc":4920},"9.news/inteligencia-artificial-aprender-cerebro-humano","UPFIaoXonc99vw41WnfvRims_9VXO8-CuD8ZutNizQM",{"id":4926,"title":4927,"author":4928,"body":4931,"description":4945,"extension":923,"meta":4946,"navigation":931,"path":4951,"seo":4952,"sitemap":4953,"stem":4954,"__hash__":4955},"content/9.news/la-poste-partners-with-pathway-to-create-digital-twin-of-fleet.md","La Poste partners with Pathway to create digital twin of fleet",{"name":4929,"img":4930,"provider":10},"IoT Insider","/assets/blog/avatars/iot-insider-av.png",{"type":12,"value":4932,"toc":4943},[4933,4935,4941],[2612,4934,2615],{"id":2614},[20,4936,2618,4937,2623],{},[717,4938,4939],{"href":4939,"rel":4940},"https://www.iotinsider.com/news/la-poste-partners-with-pathway-to-create-digital-twin-of-fleet/",[733],[2625,4942],{"url":4939},{"title":201,"searchDepth":510,"depth":510,"links":4944},[],"French postal service company, La Poste, has partnered with Pathway, to create a digital twin of the La Poste logistics fleet.",{"redirection":931,"thumbnail":4947,"tags":4949,"date":4950,"related":951,"hidden":931},{"src":4948},"https://www.iotinsider.com/wp-content/uploads/2025/06/Pathway-and-La-Poste-770x433.png",[1112],"2025-06-22","/news/la-poste-partners-with-pathway-to-create-digital-twin-of-fleet",{"title":4927,"description":4945},{"loc":4951},"9.news/la-poste-partners-with-pathway-to-create-digital-twin-of-fleet","uONdBHKd1rACi_xonfCcrhIRUuRRQUe0kzRd1J5Ez7Y",{"id":4957,"title":4958,"author":4959,"body":4962,"description":4976,"extension":923,"meta":4977,"navigation":931,"path":4982,"seo":4983,"sitemap":4984,"stem":4985,"__hash__":4986},"content/9.news/new-ai-research-claims-to-be-getting-closer-to-modeling-human-brain.md","New AI research claims to be getting closer to modeling human brain",{"name":4960,"favicon":4961,"website":4961},"Semafor","semafor.com",{"type":12,"value":4963,"toc":4974},[4964,4966,4972],[2612,4965,2615],{"id":2614},[20,4967,2618,4968,2623],{},[717,4969,4970],{"href":4970,"rel":4971},"https://www.semafor.com/article/10/01/2025/new-ai-research-claims-to-be-getting-closer-to-modeling-human-brain",[733],[2625,4973],{"url":4970},{"title":201,"searchDepth":510,"depth":510,"links":4975},[],"A new paper published by startup Pathway says it can more closely replicate connections between neurons than current LLMs",{"redirection":931,"thumbnail":4978,"tags":4980,"date":4981,"related":951},{"src":4979,"provider":10},"/assets/blog/thumbnails/new-ai-research-claims-to-be-getting-closer-to-modeling-human-brain-th.png",[1112,1830],"2025-10-01","/news/new-ai-research-claims-to-be-getting-closer-to-modeling-human-brain",{"title":4958,"description":4976},{"loc":4982},"9.news/new-ai-research-claims-to-be-getting-closer-to-modeling-human-brain","5W-pQWKhHD-8DxCgb7yIh_YxO1rdrux4ZtyJLwIxgIk",{"id":4988,"title":4989,"author":4990,"body":4993,"description":5007,"extension":923,"meta":5008,"navigation":931,"path":5013,"seo":5014,"sitemap":5015,"stem":5016,"__hash__":5017},"content/9.news/new-dragon-hatchling-ai-architecture-modeled-after-the-human-brain-could-be-a-key-step-toward-agi-researchers-claim.md","New 'Dragon Hatchling' AI architecture modeled after the human brain could be a key step toward AGI, researchers claim",{"name":4991,"favicon":4992,"website":4992},"Live Science","livescience.com",{"type":12,"value":4994,"toc":5005},[4995,4997,5003],[2612,4996,2615],{"id":2614},[20,4998,2618,4999,2623],{},[717,5000,5001],{"href":5001,"rel":5002},"https://www.livescience.com/technology/artificial-intelligence/new-dragon-hatchling-ai-architecture-modeled-after-the-human-brain-could-be-a-key-step-toward-agi-researchers-claim",[733],[2625,5004],{"url":5001},{"title":201,"searchDepth":510,"depth":510,"links":5006},[],"Scientists say a new kind of AI could bridge the gap between current systems and machines that learn and think more like us.",{"redirection":931,"thumbnail":5009,"tags":5011,"date":5012,"related":951},{"src":5010},"https://cdn.mos.cms.futurecdn.net/txftSjJw9qMtxWy85qzWFY-650-80.png.webp",[1112,1830],"2025-11-13","/news/new-dragon-hatchling-ai-architecture-modeled-after-the-human-brain-could-be-a-key-step-toward-agi-researchers-claim",{"title":4989,"description":5007},{"loc":5013},"9.news/new-dragon-hatchling-ai-architecture-modeled-after-the-human-brain-could-be-a-key-step-toward-agi-researchers-claim","8jGm5GWJ04de_gyDTYNKXLjZoOcwFgC_rLOIcioFStI",{"id":5019,"title":5020,"author":5021,"body":5022,"description":5036,"extension":923,"meta":5037,"navigation":931,"path":5041,"seo":5042,"sitemap":5043,"stem":5044,"__hash__":5045},"content/9.news/new-podcast-europes-ai-opportunity-brand.md","New podcast: Europe’s AI opportunity",{"name":2606,"img":4492},{"type":12,"value":5023,"toc":5034},[5024,5026,5032],[2612,5025,2615],{"id":2614},[20,5027,2618,5028,2623],{},[717,5029,5030],{"href":5030,"rel":5031},"https://sifted.eu/articles/new-podcast-europes-ai-opportunity-brnd",[733],[2625,5033],{"url":5030},{"title":201,"searchDepth":510,"depth":510,"links":5035},[],"Europe’s AI Opportunity is a podcast mini-series, in collaboration with Nebius, to explore the state of play for the continent’s AI startups",{"layout":925,"redirection":931,"date":1799,"thumbnail":5038,"tags":5040,"hidden":931},{"src":5039,"provider":10},"/assets/blog/thumbnails/new-podcast-europes-ai-opportunity-brand-th.png",[1112,1239],"/news/new-podcast-europes-ai-opportunity-brand",{"title":5020,"description":5036},{"loc":5041},"9.news/new-podcast-europes-ai-opportunity-brand","9skf0StJJmAOF_VlfwQU4PxvOVkY9-xsWxxyHL6DnTc",{"id":5047,"title":5048,"author":5049,"body":5050,"description":5061,"extension":923,"meta":5062,"navigation":931,"path":5069,"seo":5070,"sitemap":5071,"stem":5072,"__hash__":5073},"content/9.news/newsletter-2023-01-24.md","Pathway - a message from the CEO",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5051,"toc":5059},[5052],[5053,5054],"iframe",{"src":5055,"width":5056,"height":5056,"className":5057},"https://mailchi.mp/pathway/a_message_from_the_ceo_january23","100%",[5058],"h-screen",{"title":201,"searchDepth":510,"depth":510,"links":5060},[],"Pathway - the data processing framework which takes care of streaming data updates for you - announces a $4.5M round, and opens to developers.",{"thumbnail":5063,"date":5065,"tags":5066,"related":5068,"aside":951,"hidden":931},{"src":5064,"provider":10},"/assets/blog/thumbnails/pathway-newsletter-th.png","2023-01-24",[5067],"newsletter",[],"/news/newsletter-2023-01-24",{"title":5048,"description":5061},{"loc":5069},"9.news/newsletter-2023-01-24","7DAvhrBmy9y1MnaIpE43er3DnjE5hBpdacIKmirmyHk",{"id":5075,"title":5076,"author":5077,"body":5078,"description":5086,"extension":923,"meta":5087,"navigation":931,"path":5092,"seo":5093,"sitemap":5094,"stem":5095,"__hash__":5096},"content/9.news/newsletter-2023-05-26.md","Pathway is the most powerful: benchmarks & real-time LLMs",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5079,"toc":5084},[5080],[5053,5081],{"src":5082,"width":5056,"height":5056,"className":5083},"https://mailchi.mp/pathway/a_message_from_the_ceo_january23-15529062",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5085},[],"The fastest data processing engine on the market",{"thumbnail":5088,"date":5089,"tags":5090,"related":5091,"aside":951,"hidden":931},{"src":5064,"provider":10},"2023-05-26",[5067],[],"/news/newsletter-2023-05-26",{"title":5076,"description":5086},{"loc":5092},"9.news/newsletter-2023-05-26","Xp4h7XgzvHueutI2vQcCrFuCB0ikJByTMOfbt6p_fmc",{"id":5098,"title":5099,"author":5100,"body":5101,"description":5109,"extension":923,"meta":5110,"navigation":931,"path":5114,"seo":5115,"sitemap":5116,"stem":5117,"__hash__":5118},"content/9.news/newsletter-2023-07-25.md","Pathway: Gartner, Redpanda, Intel and more",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5102,"toc":5107},[5103],[5053,5104],{"src":5105,"width":5056,"height":5056,"className":5106},"https://mailchi.mp/pathway/a_message_from_the_ceo_january23-15524410",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5108},[],"Pathway newsletter: Gartner, Redpanda, Intel and more!",{"thumbnail":5111,"date":4039,"tags":5112,"related":5113,"aside":951,"hidden":931},{"src":5064,"provider":10},[5067],[],"/news/newsletter-2023-07-25",{"title":5099,"description":5109},{"loc":5114},"9.news/newsletter-2023-07-25","4pb7f4q1LYK6HbhgwGqxZ6C8-W3RuzUOYjukg-GGGao",{"id":5120,"title":5121,"author":5122,"body":5123,"description":5131,"extension":923,"meta":5132,"navigation":931,"path":5137,"seo":5138,"sitemap":5139,"stem":5140,"__hash__":5141},"content/9.news/newsletter-2023-11-30.md","Streaming meets AI: Live Jupyter dashboards, LLM alerts, and more",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5124,"toc":5129},[5125],[5053,5126],{"src":5127,"width":5056,"height":5056,"className":5128},"https://mailchi.mp/pathway/a_message_from_the_ceo_january23-15538130",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5130},[],"Pathway is the easiest framework for the AI era, combining live data sources such as Kafka with capacity for machine-learning powered transformation",{"thumbnail":5133,"date":5134,"tags":5135,"related":5136,"aside":951,"hidden":931},{"src":5064,"provider":10},"2023-11-30",[5067],[],"/news/newsletter-2023-11-30",{"title":5121,"description":5131},{"loc":5137},"9.news/newsletter-2023-11-30","qwP-D-D00C2u8E8FmlAEba0KEJj8lJDsHJ3Jm7hF6zk",{"id":5143,"title":5144,"author":5145,"body":5146,"description":5154,"extension":923,"meta":5155,"navigation":931,"path":5160,"seo":5161,"sitemap":5162,"stem":5163,"__hash__":5164},"content/9.news/newsletter-2024-02-15.md","Preparing for 2024 Olympics | Real-time RAG: A Spreadsheet is All You Need | ETL in Kafka",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5147,"toc":5152},[5148],[5053,5149],{"src":5150,"width":5056,"height":5056,"className":5151},"https://mailchi.mp/pathway/starting_2024_with_pathway-15544274",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5153},[],"Pathway Powers La Poste for 2024 Olympics | Real-time RAG: A Spreadsheet is All You Need",{"thumbnail":5156,"date":5157,"tags":5158,"related":5159,"aside":951,"hidden":931},{"src":5064,"provider":10},"2024-02-15",[5067],[],"/news/newsletter-2024-02-15",{"title":5144,"description":5154},{"loc":5160},"9.news/newsletter-2024-02-15","zybcJIFYmmhloOo-By6KMggIyr24HBz_BOJE7mB0Odo",{"id":5166,"title":5167,"author":5168,"body":5169,"description":5177,"extension":923,"meta":5178,"navigation":931,"path":5183,"seo":5184,"sitemap":5185,"stem":5186,"__hash__":5187},"content/9.news/newsletter-2024-04-12.md","4x LLM token cost reduction, Secure local RAG applications, and Bay Area Meetup",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5170,"toc":5175},[5171],[5053,5172],{"src":5173,"width":5056,"height":5056,"className":5174},"https://mailchi.mp/pathway/a_message_from_the_ceo_january23-17967537",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5176},[],"We will help you get your RAG App up and running in 1 week.",{"single":931,"aside":951,"thumbnail":5179,"date":5180,"tags":5181,"related":5182,"hidden":931},{"src":5064,"provider":10},"2024-04-12",[5067],[],"/news/newsletter-2024-04-12",{"title":5167,"description":5177},{"loc":5183},"9.news/newsletter-2024-04-12","0WrCAzhaGusDisnNQ8YIcuOvVPImqbrk47Y_6fLyIf8",{"id":5189,"title":5190,"author":5191,"body":5192,"description":5200,"extension":923,"meta":5201,"navigation":931,"path":5205,"seo":5206,"sitemap":5207,"stem":5208,"__hash__":5209},"content/9.news/newsletter-2025-02-25.md","Pathway to the Silicon Valley",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5193,"toc":5198},[5194],[5053,5195],{"src":5196,"width":5056,"height":5056,"className":5197},"https://mailchi.mp/pathway/newsletterpaloalto",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5199},[],"New Silicon Valley Office: Let’s Talk AI",{"thumbnail":5202,"date":5203,"tags":5204,"aside":951},{"src":5064,"provider":10},"2025-02-25",[5067],"/news/newsletter-2025-02-25",{"title":5190,"description":5200},{"loc":5205},"9.news/newsletter-2025-02-25","Ii3OAF-NcT_M_llE1hW-qYUwTumyfEWppimSMh3solM",{"id":5211,"title":5212,"author":5213,"body":5214,"description":5222,"extension":923,"meta":5223,"navigation":931,"path":5227,"seo":5228,"sitemap":5229,"stem":5230,"__hash__":5231},"content/9.news/newsletter-2025-10-15 copy.md","The next Transformer moment for AI - read in Forbes",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5215,"toc":5220},[5216],[5053,5217],{"src":5218,"width":5056,"height":5056,"className":5219},"https://mailchi.mp/pathway/the-next-transformer-moment-for-ai",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5221},[],"Our groundbreaking post-transformer BDH architecture has launched",{"thumbnail":5224,"date":5225,"tags":5226,"aside":951},{"src":5064,"provider":10},"2025-10-15",[5067],"/news/newsletter-2025-10-15-copy",{"title":5212,"description":5222},{"loc":5227},"9.news/newsletter-2025-10-15 copy","aDtKV4vTz8z4nRXDwk01AtJHFyBqMa-kmKQb8r_-T1U",{"id":5233,"title":5234,"author":5235,"body":5236,"description":5244,"extension":923,"meta":5245,"navigation":931,"path":5249,"seo":5250,"sitemap":5251,"stem":5252,"__hash__":5253},"content/9.news/newsletter-2026-01-15.md","WSJ: Pathway marks the beginning of the post-transformer era",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5237,"toc":5242},[5238],[5053,5239],{"src":5240,"width":5056,"height":5056,"className":5241},"https://mailchi.mp/pathway/the-next-transformer-moment-for-ai-17996970",[5058],{"title":201,"searchDepth":510,"depth":510,"links":5243},[],"Thank you for your support, trust and partnership in building this next foundation together",{"thumbnail":5246,"date":5247,"tags":5248,"aside":951},{"src":5064,"provider":10},"2026-01-15",[5067],"/news/newsletter-2026-01-15",{"title":5234,"description":5244},{"loc":5249},"9.news/newsletter-2026-01-15","CLHeLEpdMc70hqWtNMW-5AlIelPTJycCwWQOh-Sp8hY",{"id":5255,"title":5256,"author":5257,"body":5260,"description":201,"extension":923,"meta":5274,"navigation":931,"path":5279,"seo":5280,"sitemap":5281,"stem":5282,"__hash__":5283},"content/9.news/open-ai-coding-jobs-silicon-valley-google.md","OpenAI claims AI is making coding jobs better, not worse. Is it true?",{"name":5258,"favicon":5259,"website":5259},"Fast Company","fastcompany.com",{"type":12,"value":5261,"toc":5272},[5262,5264,5270],[2612,5263,2615],{"id":2614},[20,5265,2618,5266,2623],{},[717,5267,5268],{"href":5268,"rel":5269},"https://www.fastcompany.com/91411004/open-ai-coding-jobs-silicon-valley-google",[733],[2625,5271],{"url":5268},{"title":201,"searchDepth":510,"depth":510,"links":5273},[],{"layout":925,"redirection":931,"tags":5275,"date":5276,"thumbnail":5277},[1112],"2025-09-26",{"src":5278,"provider":10},"/assets/blog/thumbnails/fastcompany-th.png","/news/open-ai-coding-jobs-silicon-valley-google",{"title":5256,"description":201},{"loc":5279},"9.news/open-ai-coding-jobs-silicon-valley-google","0tUzqMv2fLoiiJeiSzxe4-gj2OrMOOG0TZGZL8tsXYo",{"id":5285,"title":5286,"author":5287,"body":5290,"description":5304,"extension":923,"meta":5305,"navigation":931,"path":5310,"seo":5311,"sitemap":5312,"stem":5313,"__hash__":5314},"content/9.news/opinion-eu-could-be-epicenter-of-ai-academia-as-us-cuts-funding.md","Opinion: EU could be epicenter of AI academia as US cuts funding",{"name":5288,"img":5289,"provider":10},"SiliconRepublic.com","/assets/blog/avatars/siliconrepublic-av.png",{"type":12,"value":5291,"toc":5302},[5292,5294,5300],[2612,5293,2615],{"id":2614},[20,5295,2618,5296,2623],{},[717,5297,5298],{"href":5298,"rel":5299},"https://www.siliconrepublic.com/innovation/opinion-eu-could-be-epicentre-of-ai-academia-as-us-cuts-funding",[733],[2625,5301],{"url":5298},{"title":201,"searchDepth":510,"depth":510,"links":5303},[],"EU academic research could be the big winner as US cuts funding to its research bodies",{"redirection":931,"thumbnail":5306,"tags":5308,"date":5309,"related":951,"hidden":931},{"src":5307},"https://www.siliconrepublic.com/wp-content/uploads/2025/05/Zuzanna-Stamirowska.jpg",[1112],"2025-05-16","/news/opinion-eu-could-be-epicenter-of-ai-academia-as-us-cuts-funding",{"title":5286,"description":5304},{"loc":5310},"9.news/opinion-eu-could-be-epicenter-of-ai-academia-as-us-cuts-funding","ELouOZ5nJLGrugUxdHjzWo9zXcc5GO1Zo87QvJTQ3lw",{"id":5316,"title":5317,"author":5318,"body":5321,"description":5335,"extension":923,"meta":5336,"navigation":931,"path":5341,"seo":5342,"sitemap":5343,"stem":5344,"__hash__":5345},"content/9.news/palo-alto-ai-firm-pathway-unveils-post-transformer-architecture-for-autonomous-ai.md","Palo Alto AI Firm Pathway Unveils Post-Transformer Architecture for Autonomous AI",{"name":5319,"favicon":5320,"website":5320},"Quantum Zeitgeist","quantumzeitgeist.com",{"type":12,"value":5322,"toc":5333},[5323,5325,5331],[2612,5324,2615],{"id":2614},[20,5326,2618,5327,2623],{},[717,5328,5329],{"href":5329,"rel":5330},"https://quantumzeitgeist.com/palo-alto-ai-firm-pathway-unveils-post-transformer-architecture-for-autonomous-ai/",[733],[2625,5332],{"url":5329},{"title":201,"searchDepth":510,"depth":510,"links":5334},[],"Artificial intelligence has long been celebrated for its uncanny ability to recognize patterns in vast datasets, yet it remains shackled by a static view of the world...",{"redirection":931,"thumbnail":5337,"tags":5339,"date":5340,"related":951},{"src":5338},"https://quantumzeitgeist.com/wp-content/uploads/Pathway_Image.gif",[1112,1830],"2025-08-03","/news/palo-alto-ai-firm-pathway-unveils-post-transformer-architecture-for-autonomous-ai",{"title":5317,"description":5335},{"loc":5341},"9.news/palo-alto-ai-firm-pathway-unveils-post-transformer-architecture-for-autonomous-ai","WQeDlBSH3cB5p2RNG9D5Vy2VY9UeKCzeaY_PSlbKe48",{"id":5347,"title":5348,"author":5349,"body":5350,"description":5385,"extension":923,"meta":5386,"navigation":931,"path":5391,"seo":5392,"sitemap":5393,"stem":5394,"__hash__":5395},"content/9.news/pathway-at-the-ai-action-summit.md","Pathway at the AI Action Summit! ",{"id":2644,"url":2645,"name":2646,"description":2647,"img":2648,"provider":10,"linkedin":2649},{"type":12,"value":5351,"toc":5383},[5352,5361,5364,5373,5376],[20,5353,5354,5355,5360],{},"Pathway was present at the ",[717,5356,5359],{"href":5357,"rel":5358},"https://www.elysee.fr/en/sommet-pour-l-action-sur-l-ia",[733],"AI Action Summit",". For almost a week, from February 6 to 11, 2025, Paris hosted numerous events aimed at strengthening international action towards artificial intelligence serving the general interest. The week was punctuated by scientific days, a cultural weekend, an international summit at the Grand Palais, numerous parallel events, and a day for businesses.",[20,5362,5363],{},"The AI Action Summit will be viewed as a success for the Macron presidency. As an attendee, it was a clear “brand-making event” for France with less focus on the broader AI landscape.",[20,5365,5366,5367,5372],{},"The prominent announcement was of course the Stargate-level investment in AI infrastructure for the country (",[717,5368,5371],{"href":5369,"rel":5370},"https://www.ft.com/content/fc6a2d7a-5ed6-436e-84a5-dda86fc258d3",[733],"Macron unveils plan for €109bn of AI investment in France, Feb 10","). While this investment is a good show, we need to see if the amounts announced will actually be invested and where exactly the investments will go – I am somewhat dubious.",[20,5374,5375],{},"The biggest concern with the investment is the gap in regard to investment from private French players, which is only €3bn out of over €100bn. With such low investment from French participants, questions are raised about whether the funding will actually benefit companies in the country, or if there will be billions poured into chips and data centers, largely to the profit of US and Asia-based firms.",[20,5377,5378,5380,5381],{},[36,5379,2646],{},", COO and Co-founder of French data and AI startup, ",[717,5382,3285],{"href":3284},{"title":201,"searchDepth":510,"depth":510,"links":5384},[],"Pathway was present at the AI Action Summit",{"layout":925,"thumbnail":5387,"tags":5389,"date":5390,"hidden":931},{"src":5388,"provider":10},"/assets/blog/thumbnails/pathway-at-the-ai-action-summit-th.png",[1112],"2025-02-12","/news/pathway-at-the-ai-action-summit",{"title":5348,"description":5385},{"loc":5391},"9.news/pathway-at-the-ai-action-summit","Cuf6qOwn5hHZRIO3gDwIz24zmI1X3alaM39C8lrirQE",{"id":5397,"title":5398,"author":5399,"body":5402,"description":201,"extension":923,"meta":5416,"navigation":931,"path":5421,"seo":5422,"sitemap":5423,"stem":5424,"__hash__":5425},"content/9.news/pathway-bdh-brain-inspired-ai-architecture.md","Brain-inspired AI model 'BDH' may surpass the limits of Transformers",{"name":5400,"favicon":5401,"website":5401},"Radical Data Science","xenospectrum.com",{"type":12,"value":5403,"toc":5414},[5404,5406,5412],[2612,5405,2615],{"id":2614},[20,5407,2618,5408,2623],{},[717,5409,5410],{"href":5410,"rel":5411},"https://xenospectrum.com/pathway-bdh-brain-inspired-ai-architecture/",[733],[2625,5413],{"url":5410},{"title":201,"searchDepth":510,"depth":510,"links":5415},[],{"redirection":931,"lang":5417,"tags":5418,"date":4981,"related":951,"thumbnail":5419},"japanese",[1112,1830],{"src":5420,"provider":10},"/assets/blog/thumbnails/bdh-brain-th.png","/news/pathway-bdh-brain-inspired-ai-architecture",{"title":5398,"description":201},{"loc":5421},"9.news/pathway-bdh-brain-inspired-ai-architecture","anhqPMYzT_YDg9QuMSusksSNP0QA3ji7_a-N-I55N2I",{"id":5427,"title":5428,"author":5429,"body":5431,"description":5445,"extension":923,"meta":5446,"navigation":931,"path":5450,"seo":5451,"sitemap":5452,"stem":5453,"__hash__":5454},"content/9.news/pathway-launches-a-new-post-transformer-architecture-that-paves-the-way-for-autonomous-ai.md","Pathway Launches a New “Post-Transformer” Architecture That Paves the Way for Autonomous AI",{"name":5400,"favicon":5430,"website":5430},"radicaldatascience.wordpress.com",{"type":12,"value":5432,"toc":5443},[5433,5435,5441],[2612,5434,2615],{"id":2614},[20,5436,2618,5437,2623],{},[717,5438,5439],{"href":5439,"rel":5440},"https://radicaldatascience.wordpress.com/2025/10/01/pathway-launches-a-new-post-transformer-architecture-that-paves-the-way-for-autonomous-ai/",[733],[2625,5442],{"url":5439},{"title":201,"searchDepth":510,"depth":510,"links":5444},[],"Pathway, the data company building live AI that thinks in real-time like humans do, is today introducing Baby Dragon Hatchling (BDH), a new “post-Transformer” architecture that addresses one of the most significant barriers to autonomous artificial intelligence (AI): the inability to generalize over time.",{"redirection":931,"tags":5447,"date":4981,"related":951,"thumbnail":5448},[1112,1830],{"src":5449,"provider":10},"/assets/blog/thumbnails/radicaldatascience-th.png","/news/pathway-launches-a-new-post-transformer-architecture-that-paves-the-way-for-autonomous-ai",{"title":5428,"description":5445},{"loc":5450},"9.news/pathway-launches-a-new-post-transformer-architecture-that-paves-the-way-for-autonomous-ai","HGMgNwURx8yPCNcLMG1bpH528qBTicwS-SRIU67sxEE",{"id":5456,"title":5457,"author":5458,"body":5461,"description":201,"extension":923,"meta":5475,"navigation":931,"path":5480,"seo":5481,"sitemap":5482,"stem":5483,"__hash__":5484},"content/9.news/pathway-launches-new-post-transformer-architecture-paving-the-way-for-autonomous-ai.md","Pathway launches new post-transformer architecture paving the way for autonomous AI",{"name":5459,"favicon":5460,"website":5460},"Intelligent CIO","intelligentcio.com",{"type":12,"value":5462,"toc":5473},[5463,5465,5471],[2612,5464,2615],{"id":2614},[20,5466,2618,5467,2623],{},[717,5468,5469],{"href":5469,"rel":5470},"https://www.intelligentcio.com/eu/2025/10/03/pathway-launches-new-post-transformer-architecture-paving-the-way-for-autonomous-ai/",[733],[2625,5472],{"url":5469},{"title":201,"searchDepth":510,"depth":510,"links":5474},[],{"layout":925,"redirection":931,"tags":5476,"date":5477,"thumbnail":5478},[1112,1830],"2025-10-03",{"src":5479,"provider":10},"/assets/blog/thumbnails/cio-th.png","/news/pathway-launches-new-post-transformer-architecture-paving-the-way-for-autonomous-ai",{"title":5457,"description":201},{"loc":5480},"9.news/pathway-launches-new-post-transformer-architecture-paving-the-way-for-autonomous-ai","FDhfWdV94SJDDzqsN_sy0nVHuYcu9lLIxbR24fszV4E",{"id":5486,"title":5487,"author":5488,"body":5491,"description":5505,"extension":923,"meta":5506,"navigation":931,"path":5510,"seo":5511,"sitemap":5512,"stem":5513,"__hash__":5514},"content/9.news/pathway-to-deliver-new-class-of-adaptive-and-continuously-learning-ai-systems-with-aws-and-nvidia-technologies.md","Pathway to Deliver New Class of Adaptive and Continuously Learning AI Systems with AWS and NVIDIA Technologies",{"name":5489,"favicon":5490,"website":5490},"businesswire","businesswire.com",{"type":12,"value":5492,"toc":5503},[5493,5495,5501],[2612,5494,2615],{"id":2614},[20,5496,2618,5497,2623],{},[717,5498,5499],{"href":5499,"rel":5500},"https://www.businesswire.com/news/home/20251201914013/en/Pathway-to-Deliver-New-Class-of-Adaptive-and-Continuously-Learning-AI-Systems-with-AWS-and-NVIDIA-Technologies",[733],[2625,5502],{"url":5499},{"title":201,"searchDepth":510,"depth":510,"links":5504},[],"Pathway, the data company building live AI that thinks in real-time like humans do, today announced that its groundbreaking post-Transformer BDH",{"redirection":931,"thumbnail":5507,"tags":5509,"date":4636,"related":951},{"src":5508},"https://mms.businesswire.com/media/20251201914013/en/2654091/22/pathway-logo-black.jpg",[1112,1830],"/news/pathway-to-deliver-new-class-of-adaptive-and-continuously-learning-ai-systems-with-aws-and-nvidia-technologies",{"title":5487,"description":5505},{"loc":5510},"9.news/pathway-to-deliver-new-class-of-adaptive-and-continuously-learning-ai-systems-with-aws-and-nvidia-technologies","kxT730X_OFRGdaiENT4ckYUwoRKiQndZh9rvLIDbNfg",{"id":5516,"title":5517,"author":5518,"body":5521,"description":201,"extension":923,"meta":5535,"navigation":931,"path":5540,"seo":5541,"sitemap":5542,"stem":5543,"__hash__":5544},"content/9.news/revealing-the-first-biological-ai-a-step-closer-to-singularity copy.md","Revealing the First Biological AI: A Step Closer to Singularity",{"name":5519,"img":5520,"provider":10},"This Is The World","/assets/blog/avatars/thisisworld-av.jpg",{"type":12,"value":5522,"toc":5533},[5523,5525,5531],[2612,5524,2615],{"id":2614},[20,5526,2618,5527,2623],{},[717,5528,5529],{"href":5529,"rel":5530},"https://www.youtube.com/watch?v=6_v2HG8l9oA",[733],[2625,5532],{"url":5529},{"title":201,"searchDepth":510,"depth":510,"links":5534},[],{"redirection":931,"thumbnail":5536,"tags":5538,"date":5539,"related":951},{"src":5537},"https://img.youtube.com/vi/6_v2HG8l9oA/maxresdefault.jpg",[1112,1239,1830],"2025-10-04","/news/revealing-the-first-biological-ai-a-step-closer-to-singularity-copy",{"title":5517,"description":201},{"loc":5540},"9.news/revealing-the-first-biological-ai-a-step-closer-to-singularity copy","Q0UmGNVVvIyL_F4mJw9AyduEZOvCtERvymlXZJKPgi0",{"id":5546,"title":5547,"author":5548,"body":5552,"description":201,"extension":923,"meta":5566,"navigation":931,"path":5571,"seo":5572,"sitemap":5573,"stem":5574,"__hash__":5575},"content/9.news/sds-929-dragon-hatchling-the-missing-link-between-transformers-and-the-brain-with-adrian-kosowski.md","Dragon Hatchling: The Missing Link Between Transformers and the Brain, with Adrian Kosowski (SDS 929)",{"name":5549,"img":5550,"provider":10,"website":5551},"SuperDataScience","/assets/blog/avatars/superdatascience-av.png","https://superdatascience.com",{"type":12,"value":5553,"toc":5564},[5554,5556,5562],[2612,5555,2615],{"id":2614},[20,5557,2618,5558,2623],{},[717,5559,5560],{"href":5560,"rel":5561},"https://www.superdatascience.com/podcast/sds-929-dragon-hatchling-the-missing-link-between-transformers-and-the-brain-with-adrian-kosowski",[733],[2625,5563],{"url":5560},{"title":201,"searchDepth":510,"depth":510,"links":5565},[],{"redirection":931,"thumbnail":5567,"tags":5569,"date":5570,"related":951},{"src":5568,"provider":10},"/assets/blog/thumbnails/sds-th.png",[1112,1239,1830],"2025-10-07","/news/sds-929-dragon-hatchling-the-missing-link-between-transformers-and-the-brain-with-adrian-kosowski",{"title":5547,"description":201},{"loc":5571},"9.news/sds-929-dragon-hatchling-the-missing-link-between-transformers-and-the-brain-with-adrian-kosowski","s-2qrFhzM7MmVc2baoiGgl4pNppQPwCSli5yf_6Xcao",{"id":5577,"title":5578,"author":5579,"body":5582,"description":5597,"extension":923,"meta":5598,"navigation":931,"path":5603,"seo":5604,"sitemap":5605,"stem":5606,"__hash__":5607},"content/9.news/second-most-popular-ai-paper-of-the-year-in-2025.md","Second most popular AI paper of the year in 2025",{"name":5580,"favicon":5581,"website":5581},"Hugging Face","huggingface.co",{"type":12,"value":5583,"toc":5595},[5584,5592],[20,5585,5586,5591],{},[717,5587,5590],{"href":5588,"rel":5589},"https://huggingface.co/papers/2509.26507",[733],"The Dragon Hatchling: The Missing Link between the Transformer and Models of the Brain"," ranked 2 in the Top 10 most upvoted papers on HuggingFace!",[4252,5593],{"src":5594},"https://x.com/HuggingPapers/status/2005312316829516222",{"title":201,"searchDepth":510,"depth":510,"links":5596},[],"The Dragon Hatchling: The Missing Link between the Transformer and Models of the Brain ranked 2 in the Top 10 most upvoted papers on HuggingFace!",{"layout":925,"tags":5599,"date":5600,"thumbnail":5601},[1112,1830],"2025-12-28",{"src":5602,"provider":10},"/assets/blog/thumbnails/second-most-popular-ai-paper-of-the-year-in-2025-th.jpg","/news/second-most-popular-ai-paper-of-the-year-in-2025",{"title":5578,"description":5597},{"loc":5603},"9.news/second-most-popular-ai-paper-of-the-year-in-2025","MtOdodFzGtcNnHMZOX6vCi0B7M4EeZaL2yHHaacDYks",{"id":5609,"title":5610,"author":5611,"body":5612,"description":5626,"extension":923,"meta":5627,"navigation":931,"path":5632,"seo":5633,"sitemap":5634,"stem":5635,"__hash__":5636},"content/9.news/tech-predictions-2026.md","Tech That Will Change Your Life in 2026",{"name":4614,"img":4615,"provider":10,"website":4616},{"type":12,"value":5613,"toc":5624},[5614,5616,5622],[2612,5615,2615],{"id":2614},[20,5617,2618,5618,2623],{},[717,5619,5620],{"href":5620,"rel":5621},"https://www.wsj.com/tech/ai/tech-predictions-2026-6884d6b0",[733],[2625,5623],{"url":5620},{"title":201,"searchDepth":510,"depth":510,"links":5625},[],"Folding iPhones, home robots, mind-reading tech and EV supercars are all heading your way—along with AI-induced challenges in healthcare and cybersecurity",{"redirection":931,"thumbnail":5628,"tags":5630,"date":5631,"related":951},{"src":5629},"https://images.wsj.net/im-07951141",[1112],"2025-12-26","/news/tech-predictions-2026",{"title":5610,"description":5626},{"loc":5632},"9.news/tech-predictions-2026","zcrriogVSZqpygyVsvzRKJYvh7Q9-q11_c_9K0xUawQ",{"id":5638,"title":5639,"author":5640,"body":5643,"description":5657,"extension":923,"meta":5658,"navigation":931,"path":5663,"seo":5664,"sitemap":5665,"stem":5666,"__hash__":5667},"content/9.news/that-hint-where-ai-is-heading.md","That Hint Where AI Is Heading",{"name":5641,"favicon":5642,"website":5642},"Turing Post","turingpost.com",{"type":12,"value":5644,"toc":5655},[5645,5647,5653],[2612,5646,2615],{"id":2614},[20,5648,2618,5649,2623],{},[717,5650,5651],{"href":5651,"rel":5652},"https://www.turingpost.com/p/fod133",[733],[2625,5654],{"url":5651},{"title":201,"searchDepth":510,"depth":510,"links":5656},[],"Research Papers That Hint Where AI Is Heading with commentary from their authors + foundational books to read this holiday season",{"redirection":931,"thumbnail":5659,"tags":5661,"date":5662,"related":951},{"src":5660},"https://beehiiv-images-production.s3.amazonaws.com/uploads/asset/file/644e5fdd-ea96-4dcf-b286-6783c793e66b/Frame_328.png?t=1767048275",[1112,1830],"2025-12-29","/news/that-hint-where-ai-is-heading",{"title":5639,"description":5657},{"loc":5663},"9.news/that-hint-where-ai-is-heading","30G9CIiNUUUHYjfZAQjHHGvjaHQIKc_udisd7jiyvV0",{"id":5669,"title":5590,"author":5670,"body":5671,"description":5684,"extension":923,"meta":5685,"navigation":931,"path":5691,"seo":5692,"sitemap":5693,"stem":5694,"__hash__":5695},"content/9.news/the-dragon-hatchling-the-missing-link-between-the-transformer-and-models-of-the-brain.md",{"name":5580,"favicon":5581,"website":5581},{"type":12,"value":5672,"toc":5682},[5673,5675,5680],[2612,5674,2615],{"id":2614},[20,5676,2618,5677,2623],{},[717,5678,5588],{"href":5588,"rel":5679},[733],[2625,5681],{"url":5588},{"title":201,"searchDepth":510,"depth":510,"links":5683},[],"BDH, a biologically inspired Large Language Model, combines scale-free network architecture with Hebbian learning to achieve Transformer-like performance while maintaining interpretability.",{"redirection":931,"thumbnail":5686,"tags":5688,"date":5690,"related":951},{"src":5687,"provider":10},"/assets/blog/thumbnails/hugging-face-th.png",[1112,1830,5689],"developer","2025-09-30","/news/the-dragon-hatchling-the-missing-link-between-the-transformer-and-models-of-the-brain",{"title":5590,"description":5684},{"loc":5691},"9.news/the-dragon-hatchling-the-missing-link-between-the-transformer-and-models-of-the-brain","Flg7qhuKdidFidYKH_gkSUYbCxLzU7Tw4XW-eSSh_aw",{"id":5697,"title":5698,"author":5699,"body":5700,"description":5714,"extension":923,"meta":5715,"navigation":931,"path":5720,"seo":5721,"sitemap":5722,"stem":5723,"__hash__":5724},"content/9.news/the-post-transformer-era-ais-next-frontier-nyu-x-pathway.md","The Post-Transformer Era: AI's Next Frontier | NYU x Pathway",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5701,"toc":5712},[5702,5704,5710],[2612,5703,2615],{"id":2614},[20,5705,2618,5706,2623],{},[717,5707,5708],{"href":5708,"rel":5709},"https://www.youtube.com/watch?v=o9o7fU_ZSIE",[733],[2625,5711],{"url":5708},{"title":201,"searchDepth":510,"depth":510,"links":5713},[],"What comes after Transformers? The Transformer architecture behind GPT has dominated AI for nearly a decade. But cracks are showing. Transformer-based models have no continuous learning (frozen in time, like Groundhog Day), limited context windows, and compute costs that spiral as reasoning gets longer",{"redirection":931,"thumbnail":5716,"tags":5718,"date":5719,"related":951},{"src":5717,"contain":931},"https://img.youtube.com/vi/o9o7fU_ZSIE/maxresdefault.jpg",[1239,1238,1830],"2026-02-06","/news/the-post-transformer-era-ais-next-frontier-nyu-x-pathway",{"title":5698,"description":5714},{"loc":5720},"9.news/the-post-transformer-era-ais-next-frontier-nyu-x-pathway","8iYX_k2Pt0xEqJWlRqsyL6BE2FQhbC0VoPYeFze5T7U",{"id":5726,"title":5727,"author":5728,"body":5731,"description":201,"extension":923,"meta":5745,"navigation":931,"path":5750,"seo":5751,"sitemap":5752,"stem":5753,"__hash__":5754},"content/9.news/this-ai-grows-a-brain-during-training.md","This AI Grows a Brain During Training (Pathway's AI w/ Zuzanna Stamirowska)",{"name":5729,"img":5730,"provider":10},"The Neuron","/assets/blog/avatars/the-neuron-av.jpg",{"type":12,"value":5732,"toc":5743},[5733,5735,5741],[2612,5734,2615],{"id":2614},[20,5736,2618,5737,2623],{},[717,5738,5739],{"href":5739,"rel":5740},"https://www.youtube.com/watch?v=duw7RUif8hE",[733],[2625,5742],{"url":5739},{"title":201,"searchDepth":510,"depth":510,"links":5744},[],{"redirection":931,"thumbnail":5746,"tags":5748,"date":5749,"related":951},{"src":5747,"contain":931,"provider":10},"/assets/blog/thumbnails/this-ai-grows-a-brain-during-training-th.jpg",[1112,1239,1830],"2026-01-06","/news/this-ai-grows-a-brain-during-training",{"title":5727,"description":201},{"loc":5750},"9.news/this-ai-grows-a-brain-during-training","iINZWFKKkoI_5EHwc3ne2p1qKZgsI1bJNVcQqsJjB5Q",{"id":5756,"title":5757,"author":5758,"body":5761,"description":5757,"extension":923,"meta":5775,"navigation":931,"path":5779,"seo":5780,"sitemap":5781,"stem":5782,"__hash__":5783},"content/9.news/transdev-and-pathway-partner.md","Transdev and Pathway partner to improve mobility and public transport performance through LiveAI™",{"name":5759,"img":5760},"transdev","https://media.glassdoor.com/sql/413452/transdev-squareLogo-1702746543089.png",{"type":12,"value":5762,"toc":5773},[5763,5765,5771],[2612,5764,2615],{"id":2614},[20,5766,2618,5767,2623],{},[717,5768,5769],{"href":5769,"rel":5770},"https://www.transdev.com/en/press-release/transdev-and-pathway-partner/",[733],[2625,5772],{"url":5769},{"title":201,"searchDepth":510,"depth":510,"links":5774},[],{"redirection":931,"layout":950,"aside":951,"single":931,"date":1817,"thumbnail":5776,"tags":5778,"hidden":931},{"src":5777},"https://www.transdev.com/wp-content/uploads/2025/04/bus-electrique-dkbus-a-dunkerque-1800x768.jpg",[1112,1021],"/news/transdev-and-pathway-partner",{"title":5757,"description":5757},{"loc":5779},"9.news/transdev-and-pathway-partner","9das8HOuFhBCwoIdiHVM-dc2lt23ezIy-LEC9H9kfvY",{"id":5785,"title":5757,"author":5786,"body":5787,"description":5876,"extension":923,"meta":5877,"navigation":931,"path":5881,"seo":5882,"sitemap":5883,"stem":5884,"__hash__":5885},"content/9.news/transdev-pathway-live-ai-public-transport-mobility.md",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":5788,"toc":5871},[5789,5804,5807,5810,5814,5817,5820,5828,5836,5843,5849,5852,5856,5864,5868],[20,5790,5791,5794,5795,5797,5798,5803],{},[36,5792,5793],{},"Issy-les-Moulineaux (France), April 17, 2025"," - ",[717,5796,3285],{"href":3284},", the data company that builds LiveAI™, and ",[717,5799,5802],{"href":5800,"rel":5801},"https://www.transdev.com",[733],"Transdev",", a global daily mobility solutions provider, today announce a strategic partnership to transform public transport network operations by integrating innovative real-time AI frameworks at their heart.",[20,5805,5806],{},"With its operations and systems constantly generating more and more data, Transdev needed a solution capable of processing real-time information, combined with historical data, to address mobility challenges in various territories.",[20,5808,5809],{},"Pathway, which enables continuous learning via an efficient and scalable data engine to create LiveAI™ systems that think and learn in real time, allows Transdev to leverage AI systems powered by live data pipelines. This delivers insights and supports decision-making based on always up-to-date knowledge, which is crucial for the transportation sector.\nThe Relevance of AI: Transforming Simple Geolocation Data into Reli",[15,5811,5813],{"id":5812},"the-relevance-of-ai-transforming-simple-geolocation-data-into-reliable-real-time-passenger-information","The Relevance of AI: Transforming Simple Geolocation Data into Reliable Real-Time Passenger Information",[20,5815,5816],{},"For public transport applications, Pathway's LiveAI™ technology is specifically applied to geospatial and temporal data. For example, it enables live analytics of how vehicles are moving through a city in real time. Other data regarding the operation or functioning of vehicles can also be analyzed through AI.",[20,5818,5819],{},"The benefits of applying LiveAI™ to transport operations include:",[69,5821,5822,5825],{},[72,5823,5824],{},"Increased operational efficiency: More accurate and reliable predictions of arrival times, enhanced dynamic management of planned and unplanned diversions and disruptions, and new high-performance tools for operators.",[72,5826,5827],{},"Improved customer experience: Providing reliable, real-time passenger information, both in normal and disrupted situations to optimize user experience. The automated management of new arrival times when routes are diverted also reduces passenger waiting times and inconvenience.",[5829,5830,5833],"quote",{"name":5831,"title":5832},"Edouard Hénaut","CEO France at Transdev",[20,5834,5835],{},"Pathway shows that real-time data processing and AI integrate seamlessly into our chain of business tools. The solution complements our operating and passenger information systems, without adding complexity or implementation delays, while guaranteeing reliable results. Effectively ensuring daily mobility requires expert use of the real-time data we produce, transform, and deliver to both passengers and local authorities clients. Our partnership with Pathway strengthens our expertise in this area and enables significant gains for the benefit of service quality.",[5829,5837,5840],{"name":5838,"title":5839},"Laurent Mahieu","Director of the Hauts-de-France and Grand-Est Regions at Transdev, President of DK’BUS",[20,5841,5842],{},"The experimentation on the DK’BUS network, which serves the Urban Community of Dunkirk, has shown significant gains in terms of information quality. The accuracy and reliability of the information delivered has improved, enabling the disappearance of theoretical arrival times, better prediction of waiting time estimations and efficient dynamic management of unscheduled deviations. We look forward to deploying and offering this quality of information to our passengers daily. The DK’Bus team, led by General Manager Nicolas Gaillard, already has great ideas to introduce it to our travelers! Pathway is undeniably the high-performance solution for monitoring and visualizing our activity to become even more reactive and continuously improve the data we produce.",[5829,5844,5846],{"name":2846,"title":5845},"CEO and co-founder of Pathway",[20,5847,5848],{},"Transportation is a dynamic industry, and an understanding of the real-time situation is critical for navigating mobility challenges. Applying AI and the most modern data processing to these challenges makes a very tangible difference to communities. We are proud to provide AI that reduces wait times and unpredictability, in turn attracting more people to public transport and boosting quality of life for residents of cities around the world.",[20,5850,5851],{},"Pathway's technology is proven with over 46,000 installations and users and is supported through the ZEBOX ecosystem that gathers companies like CMA CGM, Transdev or VINCI.  Pathway now also delivers for major clients such as NATO and La Poste. The company recently raised $10 million in seed funding and has a growing community of developers based in over 100 countries.",[15,5853,5855],{"id":5854},"about-transdev","About Transdev",[20,5857,5858,5859],{},"Operator and leading independent private mobility group, Transdev empowers freedom to move every day thanks to safe, reliable and innovative solutions that serve the common good. Present in 19 countries, Transdev transports an average of 12.8 million passengers daily, operating all transportation modes and resolutely committed to the ecological transition. The Group employs more than 105,000 women and men serving its passengers, consolidating its position as the world leader in public transportation. Transdev advises and supports local authorities and companies in a long-term partnership. Transdev is jointly owned by Caisse des Dépôts (66%) and the Rethmann Group (34%). In 2024, Transdev reported sales of €10.05 billion. For more information: ",[717,5860,5863],{"href":5861,"rel":5862},"http://www.transdev.com",[733],"www.transdev.com",[15,5865,5867],{"id":5866},"about-pathway","About Pathway",[5869,5870],"pathway-about",{},{"title":201,"searchDepth":510,"depth":510,"links":5872},[5873,5874,5875],{"id":5812,"depth":510,"text":5813},{"id":5854,"depth":510,"text":5855},{"id":5866,"depth":510,"text":5867},"Transdev and Pathway announce a strategic partnership to revolutionize public transport using LiveAI™, enhancing real-time passenger information and operational efficiency across mobility networks",{"layout":925,"thumbnail":5878,"tags":5880,"date":4827,"hidden":931},{"src":5879,"provider":10},"/assets/blog/thumbnails/transdev-pathway-live-ai-public-transport-mobility-th.png",[1112],"/news/transdev-pathway-live-ai-public-transport-mobility",{"title":5757,"description":5876},{"loc":5881},"9.news/transdev-pathway-live-ai-public-transport-mobility","-2bRmDWuA_uvKMq9WgwncU77QQFzmUcluM7MjqQjhSw",{"id":5887,"title":5888,"author":5889,"body":5892,"description":5906,"extension":923,"meta":5907,"navigation":931,"path":5912,"seo":5913,"sitemap":5914,"stem":5915,"__hash__":5916},"content/9.news/victor-szczerba-assumes-cco-role-at-pathway-post-funding.md","Victor Szczerba assumes CCO role at Pathway post funding",{"name":5890,"img":5891,"provider":10},"ITBrief","/assets/blog/avatars/itbrief-av.png",{"type":12,"value":5893,"toc":5904},[5894,5896,5902],[2612,5895,2615],{"id":2614},[20,5897,2618,5898,2623],{},[717,5899,5900],{"href":5900,"rel":5901},"https://itbrief.news/story/victor-szczerba-assumes-cco-role-at-pathway-post-funding",[733],[2625,5903],{"url":5900},{"title":201,"searchDepth":510,"depth":510,"links":5905},[],"Victor Szczerba has been appointed Chief Commercial Officer at Pathway, following its USD $10 million funding, to enhance its LiveAI™ technology and strategies",{"redirection":931,"layout":950,"date":5908,"thumbnail":5909,"tags":5911,"hidden":931},"2025-04-01",{"src":5910,"provider":10},"/assets/blog/thumbnails/victor-szczerba-assumes-cco-role-at-pathway-post-funding-th.png",[1112],"/news/victor-szczerba-assumes-cco-role-at-pathway-post-funding",{"title":5888,"description":5906},{"loc":5912},"9.news/victor-szczerba-assumes-cco-role-at-pathway-post-funding","hueHB0Zx7yKl2eyqycaLbit82Ii2T_Ioe1fcjPCmw84",{"id":5918,"title":5919,"author":5920,"body":5923,"description":5937,"extension":923,"meta":5938,"navigation":931,"path":5943,"seo":5944,"sitemap":5945,"stem":5946,"__hash__":5947},"content/9.news/why-todays-ai-struggles-with-the-real-world-and-what-comes-next.md","Why today’s AI struggles with the real world, and what comes next",{"name":5921,"favicon":5922,"website":5922},"ET Edge Insights","etedge-insights.com",{"type":12,"value":5924,"toc":5935},[5925,5927,5933],[2612,5926,2615],{"id":2614},[20,5928,2618,5929,2623],{},[717,5930,5931],{"href":5931,"rel":5932},"https://etedge-insights.com/technology/artificial-intelligence/why-todays-ai-struggles-with-the-real-world-and-what-comes-next/",[733],[2625,5934],{"url":5931},{"title":201,"searchDepth":510,"depth":510,"links":5936},[],"Why AI's biggest challenge is memory—and how post-transformer architectures could enable continuous learning",{"redirection":931,"thumbnail":5939,"tags":5941,"date":5942,"related":951},{"src":5940},"https://etedge-insights.com/wp-content/uploads/2025/12/AI-Quantum.jpg",[1112,1830],"2026-03-12","/news/why-todays-ai-struggles-with-the-real-world-and-what-comes-next",{"title":5919,"description":5937},{"loc":5943},"9.news/why-todays-ai-struggles-with-the-real-world-and-what-comes-next","i1OIYyBg13TYXC4qX5A8Y1fpFdrsyAjtlY8tEwLcbj0",{"id":5949,"title":5950,"author":5951,"body":5954,"description":201,"extension":923,"meta":5968,"navigation":931,"path":5972,"seo":5973,"sitemap":5974,"stem":5975,"__hash__":5976},"content/9.news/zuzanna-stamirowska-co-founder-and-ceo-of-pathway-interview-series.md","Zuzanna Stamirowska, Co-Founder and CEO of Pathway – Interview Series",{"name":5952,"favicon":5953,"website":5953},"Unite AI","unite.ai",{"type":12,"value":5955,"toc":5966},[5956,5958,5964],[2612,5957,2615],{"id":2614},[20,5959,2618,5960,2623],{},[717,5961,5962],{"href":5962,"rel":5963},"https://www.unite.ai/zuzanna-stamirowska-co-founder-and-ceo-of-pathway-interview-series/",[733],[2625,5965],{"url":5962},{"title":201,"searchDepth":510,"depth":510,"links":5967},[],{"layout":925,"redirection":931,"tags":5969,"date":5276,"thumbnail":5970},[1112],{"src":5971,"provider":10},"/assets/blog/thumbnails/zuzanna-stamirowska-co-founder-and-ceo-of-pathway-interview-series-th.png","/news/zuzanna-stamirowska-co-founder-and-ceo-of-pathway-interview-series",{"title":5950,"description":201},{"loc":5972},"9.news/zuzanna-stamirowska-co-founder-and-ceo-of-pathway-interview-series","__oquUzUBg3yWehTG8IOEd0HmtaYY207jjOBr4uJnvA",{"id":5978,"title":944,"author":2595,"body":5979,"description":6014,"extension":923,"meta":6015,"navigation":931,"path":945,"seo":6016,"sitemap":6017,"stem":946,"__hash__":6018},"content/framework/blog/1.index.md",{"type":12,"value":5980},[5981,5985,6000,6011],[2612,5982,5984],{"id":5983},"in-the-news","In the news",[20,5986,5987,5988,5993,5994,5999],{},"Read about ",[392,5989,5992],{"className":5990},[5991],"text-primary-500","Pathway’s"," latest ",[392,5995,5998],{"className":5996},[5997],"text-secondary-500","media mentions, press releases"," and more!",[6001,6002,6008],"modal",{"className":6003,"name":6007},[3527,3528,204,6004,6005,585,6006],"my-4","mb-10","text-base","Newsletter",[20,6009,6010],{},"Subscribe to our newsletter!",[6012,6013],"articles",{},"Read about Pathway’s latest media mentions, press releases and more!",{"layout":950,"aside":951,"toc":951,"single":931},{"title":944,"description":6014},{"loc":945},"shdvdCYJ6w8bdpejuMQbMXk7dIJ7X57V5A2-R5_Ixek",{"id":6020,"title":953,"author":6021,"body":6022,"description":6168,"extension":923,"meta":6169,"navigation":931,"path":954,"seo":6172,"sitemap":6173,"stem":955,"__hash__":6174},"content/framework/blog/1.pathway-open-beta-announced.md",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":6023,"toc":6164},[6024,6028,6046,6050,6053,6056,6070,6075,6078,6081,6092,6095,6102,6106,6109,6123,6128,6135,6142,6154],[2612,6025,6027],{"id":6026},"pathway-is-now-available-to-all-developers","Pathway is now available to all developers!",[20,6029,6030,6031,6037,6038,6040],{},"Pathway - the stream processing framework which takes care of data updates for you - announces a ",[717,6032,6033],{"href":4501},[6034,6035,6036],"b",{},"$4.5M funding round",", and opens to all developers. You can try it out in a cloud notebook directly from your browser, or run it on a local Linux machine.",[4438,6039],{},[717,6041,6045],{"className":6042,"bold":201,"size":6044,"href":719},[6043,3527,3528],"mb-0!","large","Run it now",[15,6047,6049],{"id":6048},"why-should-you-use-pathway","Why should you use Pathway?",[20,6051,6052],{},"Pathway is a programming framework which allows you to work with streaming data as if you were working with static data, in batch mode.",[20,6054,6055],{},"Have you ever tried to make sense of streaming data? If so, there is a high chance that you encountered at least one of these issues:",[69,6057,6058,6061,6064,6067],{},[72,6059,6060],{},"There are these annoying data updates that need to be taken care of",[72,6062,6063],{},"One needs to use the same logic to handle real-time and historical data",[72,6065,6066],{},"Debugging is a nightmare, because how can you debug something against unknown data?",[72,6068,6069],{},"Not to mention applying proper Machine Learning on top of streaming data to draw business insights from it. Business insights, which are necessary for key decision-making.",[20,6071,6072],{},[3130,6073],{"alt":201,"src":6074},"/assets/content/blog/difficulties-streaming-data.svg",[20,6076,6077],{},"If these are problems you have been up against, you are in the right place.\nAt Pathway, we design the programming framework which quietly takes care of data updates for you.",[20,6079,6080],{},"It gives you:",[69,6082,6083,6086,6089],{},[72,6084,6085],{},"A native real-time approach. Every task is either real-time streaming or streaming with historical data (backfilling), no need for batch, no hacks required.",[72,6087,6088],{},"Reactivity.",[72,6090,6091],{},"Full power of Python (to make all your ML dreams come true) with an extra SQL syntax layer coming soon (to make sure all data engineers are happy with their Pathway pipelines, too).",[20,6093,6094],{},"In the design of Pathway's streaming engine, we opted for ease-of-use and scalability.",[5829,6096,6099],{"name":6097,"title":6098},"Lukasz Kaiser","Co-author of Tensor Flow and co-inventor of Transformers, now at OpenAI - and an angel investor in Pathway.",[20,6100,6101],{},"In Machine Learning, the key to success of a programming framework is how to combine usability with scalability. This was the axis of competition between Google's TensorFlow and Facebook's PyTorch during the deep learning revolution. Today, Pathway has taken into account the lessons learned during this battle of giants, and embedded them in the compiler of its real-time data processing framework.",[15,6103,6105],{"id":6104},"what-does-all-this-mean-in-practice-for-a-developer","What does all this mean in practice, for a developer?",[20,6107,6108],{},"That you can write as if you were writing a batch data processing pipeline (well, it's actually a little more than that, as we support loops and iteration!) and have it run on streaming data.",[20,6110,6111,6112,6116,6117,6122],{},"For a start, check out this simple ",[717,6113,6115],{"href":6114},"/developers/templates/etl/lsh_chapter1","example of classification of handwritten digits",". All of it is captured by the code below. We approach this task with a classifier from Pathway's standard library, in this case, k-Nearest-Neighbors (",[717,6118,6121],{"href":6119,"rel":6120},"https://en.wikipedia.org/wiki/K-nearest_neighbors_algorithm",[733],"read more on Wikipedia","). Pathway builds up the corresponding control flow graph, and updates it in streaming mode.",[20,6124,6125],{},[3130,6126],{"alt":201,"src":6127},"/assets/content/blog/classification-control-flow-pathway.svg",[20,6129,6130,6131,3267],{},"Using Pathway means that all Machine Learning outcomes are updated as the models learn with new samples and improve over time. Classification decisions for tested elements will also be revisited whenever they change. Such an approach is called reactive processing of streaming data. If you would like to learn more about this topic, we explain it in detail in ",[717,6132,6134],{"href":6133},"/blog/pydata","this fresh video talk",[20,6136,6137,6138,3267],{},"You will also find many more examples in our Documentation - and we are also sharing with you the whole examples pack at ",[717,6139,6140],{"href":6140,"rel":6141},"https://github.com/pathwaycom/pathway-examples",[733],[20,6143,6144,6145,6148,6149,6153],{},"All of this is now open for you to play with it, test, and have fun. You can even ",[717,6146,6147],{"href":719},"run it"," in a cloud notebook from your browser, unless you prefer to ",[6150,6151,6152],"code",{},"pip install"," directly on to your own Linux machine.",[20,6155,6156,6157],{},"If you have some feedback on Pathway, or just some streaming use cases that are leaving you with sleepless nights, we would love to know.\n",[36,6158,6159,6163],{},[717,6160,6162],{"href":6161},"https://discord.com/invite/pathway","Join us on Discord"," or drop us a line!",{"title":201,"searchDepth":510,"depth":510,"links":6165},[6166,6167],{"id":6048,"depth":510,"text":6049},{"id":6104,"depth":510,"text":6105},"Pathway - the streaming programming framework which takes care of data updates for you - is now available to all developers.",{"layout":925,"thumbnail":6170,"tags":6171,"date":961,"hidden":931},{"src":958,"provider":10},[925,960],{"title":953,"description":6168},{"loc":954},"BaAkCMeb3W1SpF7tGK48g-lFSBWY8sX_iAPfNB20NoA",{"id":6176,"title":963,"author":6177,"body":6178,"description":963,"extension":923,"meta":6197,"navigation":931,"path":964,"seo":6200,"sitemap":6201,"stem":965,"__hash__":6202},"content/framework/blog/1000.Ilab2021.md",{"id":2844,"url":2845,"name":2846,"description":2847,"img":2848,"provider":10,"linkedin":2849},{"type":12,"value":6179,"toc":6195},[6180,6184,6192],[2612,6181,6183],{"id":6182},"pathway-ex-navalgo-named-2021-i-lab-laureate","Pathway (Ex-NavAlgo) named 2021 i-Lab Laureate",[20,6185,6186,6187,3267],{},"Feeling honored to share today that we are a 2021 laureate of the prestigious i-Lab contest organized by ",[717,6188,6191],{"href":6189,"rel":6190},"https://www.bpifrance.fr/",[733],"BPI France",[20,6193,6194],{},"Launched in 1999, this contest awards French most promising Deeptech startups, and has historically paved the way for success.",{"title":201,"searchDepth":510,"depth":510,"links":6196},[],{"layout":925,"aside":951,"thumbnail":6198,"tags":6199,"date":970,"hidden":931},{"src":968,"provider":10,"contain":931},[925],{"title":963,"description":963},{"loc":964},"acIeSyLHYBDc0NjAlOkyKOWpFm29wSRTSjLkT3bFHEU",{"id":6204,"title":972,"author":6205,"body":6206,"description":8144,"extension":923,"meta":8145,"navigation":931,"path":973,"seo":8149,"sitemap":8150,"stem":974,"__hash__":8151},"content/framework/blog/1001.gemini-rag.md",{"id":1440,"url":3120,"name":3121,"img":1141,"provider":10},{"type":12,"value":6207,"toc":8125},[6208,6212,6216,6227,6240,6257,6263,6267,6293,6297,6301,6307,6313,6318,6332,6336,6339,6353,6357,6368,6371,6376,6380,6394,6397,6411,6415,6421,6424,6427,6431,6438,6444,6447,6450,6454,6458,6461,6464,6468,6500,6503,6507,6513,6517,6524,6566,6570,6573,6594,6617,6820,6824,6827,7025,7029,7036,7043,7059,7153,7157,7160,7260,7264,7268,7271,7365,7369,7384,7457,7461,7464,7650,7654,7657,7734,7738,7741,7815,7819,7822,7887,7977,8047,8099,8107,8110,8112,8115,8121],[6209,6210],"true-img",{"alt":6211,"src":977},"blog banner",[2612,6213,6215],{"id":6214},"multimodal-rag-with-pathway-and-gemini","Multimodal RAG with Pathway and Gemini",[20,6217,6218,6219,6222,6223,6226],{},"The recent release of ",[36,6220,6221],{},"Google Gemini 1.5",", with its impressive ",[36,6224,6225],{},"1 million token context length window",", has sparked discussions about the future of RAG. However, it hasn't rendered it obsolete. This system still offers unique advantages, especially in curating and optimizing the context provided to the model, ensuring relevance and accuracy. What is particularly interesting is how these advancements can be harnessed to enhance our projects and streamline our workflows.",[20,6228,6229,6230,6233,6234,4130,6236,6239],{},"In this article, you'll learn how to set up a ",[36,6231,6232],{},"Multimodal Retrieval-Augmented Generation (MM-RAG)"," system using ",[36,6235,3285],{},[36,6237,6238],{},"Google Gemini",". You will walk through each step comprehensively, ensuring a solid understanding of both the theoretical and practical aspects of implementing Multimodal LLM and RAG applications.",[20,6241,6242,6243,4130,6246,6248,6249,6252,6253,3267],{},"You'll explore how to leverage the capabilities of ",[36,6244,6245],{},"Gemini 1.5 Flash",[36,6247,3285],{}," together. If you're interested in building RAG pipelines with OpenAI, we also have an article on ",[36,6250,6251],{},"Multimodal RAG using GPT-4o",", which you can check out ",[717,6254,6256],{"href":6255},"/developers/templates/rag/multimodal-rag","here",[20,6258,6259,6260,3267],{},"If you want to skip the explanations, you can directly find the code ",[717,6261,6256],{"href":6262},"#hands-on-multimodal-rag-with-google-gemini",[15,6264,6266],{"id":6265},"what-this-article-will-cover","What this article will cover:",[69,6268,6269,6272,6275,6278,6281,6284,6287,6290],{},[72,6270,6271],{},"What is Retrieval-Augmented Generation (RAG)?",[72,6273,6274],{},"Multimodality in LLMs",[72,6276,6277],{},"Why is Multimodal RAG (MM-RAG) Needed?",[72,6279,6280],{},"What is Multimodal RAG and Use Cases?",[72,6282,6283],{},"Gemini Models",[72,6285,6286],{},"Release of Gemini 1.5 and its impact on RAG architectures",[72,6288,6289],{},"Comparing LlamaIndex and Pathway",[72,6291,6292],{},"Hands-on Multimodal RAG with Google Gemini",[15,6294,6296],{"id":6295},"foundational-concepts","Foundational Concepts",[64,6298,6300],{"id":6299},"why-is-multimodal-rag-needed","Why is Multimodal Rag needed?",[20,6302,6303,6306],{},[36,6304,6305],{},"Retrieval-Augmented Generation (RAG)"," enhances large language models by incorporating external knowledge sources before generating responses. This approach ensures relevant and accurate output. In today's data-rich world, documents often combine text and images to convey information comprehensively. However, most Retrieval Augmented Generation (RAG) systems overlook the valuable insights locked within images. As Multimodal Large Language Models (LLMs) gain prominence, it's crucial to explore how we can leverage visual content alongside text in RAG, unlocking a deeper understanding of the information landscape.",[20,6308,6309,6312],{},[36,6310,6311],{},"Multimodal RAG"," is an advanced form of Retrieval-Augmented Generation (RAG) that goes beyond text to incorporate various data types like images, charts, and tables. This expanded capability allows for a deeper understanding of complex information, leading to more accurate and informative outputs.",[6314,6315,6317],"h4",{"id":6316},"two-options-for-multimodal-rag","Two options for Multimodal RAG",[268,6319,6320,6326],{},[72,6321,6322,6325],{},[36,6323,6324],{},"Multimodal Embeddings"," -\nThe multimodal embeddings model generates vectors based on the input you provide, which can include a combination of image, text, and video data. The image embedding vector and text embedding vector are in the same semantic space with the same dimensionality. Consequently, these vectors can be used interchangeably for use cases like searching image by text, or searching video by image.\nUtilize multimodal embeddings to integrate text and images, retrieve relevant content through similarity search, and then provide both the raw image and text chunks to a multimodal LLM for answer synthesis.",[72,6327,6328,6331],{},[36,6329,6330],{},"Text Embeddings"," -\nGenerate text summaries of images using a multimodal LLM, embed and retrieve the text, and then pass the text chunks to the LLM for answer synthesis.",[6314,6333,6335],{"id":6334},"comparing-text-based-and-multimodal-rag","Comparing text-based and multimodal RAG",[20,6337,6338],{},"Multimodal RAG offers several advantages over text-based RAG:",[69,6340,6341,6347],{},[72,6342,6343,6346],{},[36,6344,6345],{},"Enhanced knowledge access",": Multimodal RAG can access and process both textual and visual information, providing a richer and more comprehensive knowledge base for the LLM.",[72,6348,6349,6352],{},[36,6350,6351],{},"Improved reasoning capabilities",": By incorporating visual cues, multimodal RAG can make better informed inferences across different types of data modalities.",[6314,6354,6356],{"id":6355},"key-advantages-of-mm-rag","Key Advantages of MM-RAG:",[69,6358,6359,6362,6365],{},[72,6360,6361],{},"Comprehensive Understanding: Processes multiple data formats for a better picture.",[72,6363,6364],{},"Improved Performance: Visual data enhances efficiency in complex tasks.",[72,6366,6367],{},"Versatile Applications: Useful in finance, healthcare, scientific research, and more.",[64,6369,6283],{"id":6370},"gemini-models",[20,6372,6373,6375],{},[36,6374,988],{}," is Google's most capable and general AI model to date. Google has released several Gemini model variants, each tailored for different use cases and performance requirements.",[6314,6377,6379],{"id":6378},"main-gemini-models","Main Gemini Models:",[69,6381,6382,6385,6388,6391],{},[72,6383,6384],{},"Gemini Ultra: The most powerful and advanced model, capable of handling complex tasks and offering state-of-the-art performance.",[72,6386,6387],{},"Gemini Pro: A versatile model that balances performance and efficiency, suitable for a wide range of applications.",[72,6389,6390],{},"Gemini Advanced: Designed for a broader set of tasks, offering a good balance of capabilities.",[72,6392,6393],{},"Gemini Lite: A smaller, more efficient model focused on speed and responsiveness, ideal for resource-constrained environments.",[20,6395,6396],{},"Additional Variants:",[69,6398,6399,6402,6405,6408],{},[72,6400,6401],{},"Gemini 1.5 Flash: Optimized for high-volume, cost-effective applications.",[72,6403,6404],{},"Gemini 1.5 Pro: Offers a balance of performance and capabilities.",[72,6406,6407],{},"Gemini 1.0 Pro Vision: Includes vision capabilities for processing images and videos.",[72,6409,6410],{},"Gemini 1.0 Pro: Text-based model for general language tasks.",[6314,6412,6414],{"id":6413},"benefits-of-building-with-gemini","Benefits of Building with Gemini:",[20,6416,6417,6420],{},[36,6418,6419],{},"Free Credits",": Google Cloud offers new users up to $300 in free credits. This can be used to experiment with Gemini models and other Google Cloud services.\nYou can also seamlessly integrate MM-RAG applications with Google's Vertex AI platform for streamlined machine learning workflows.",[64,6422,6286],{"id":6423},"release-of-gemini-15-and-its-impact-on-rag-architectures",[20,6425,6426],{},"The Gemini 1.5 Flash model, released on May 24, 2024, revolutionized AI with its enhanced speed, efficiency, cost-effectiveness, long context window, and multimodal reasoning capabilities.",[6314,6428,6430],{"id":6429},"did-google-gemini-15-kill-the-need-of-rag","Did Google Gemini 1.5 Kill the need of RAG?",[20,6432,6433,6434,6437],{},"In one word ",[36,6435,6436],{},"“No”",". Gemini 1.5, with a 1M context length window, has sparked a new debate about whether RAG (Retrieval Augmented Generation) is still relevant or not. LLMs commonly struggle with hallucination. To address this challenge, two solutions were introduced, one involving an increased context window and the other utilizing RAG. Gemini 1.5 outperforms Claude 2.1 and GPT-4 Turbo as it can assimilate entire code bases, process over 100 papers, and various documents, but it surely hasn’t killed RAG.",[20,6439,6440,6441,3267],{},"RAG leverages your private knowledge database for effective Q&A while ensuring the security of sensitive information like trade secrets, confidential IP, GDPR-protected data, and internal documents. For more detailed insights explore our article on Private RAG with Connected Data Sources using Mistral, Ollama, and Pathway ",[717,6442,6256],{"href":6443},"/developers/templates/rag/private-rag-ollama-mistral",[20,6445,6446],{},"Additionally in traditional RAG pipelines, you can enhance performance by tweaking the retrieval process, changing the embedding model, adjusting chunking strategies, or improving source data. However, with a \"stuff-the-context-window-1M-tokens\" strategy, your only option is to improve the source data since all data is given to the model within the token limit. Additionally the context window may be filled with many relevant facts, but 40% or more of them are “lost” to the model. If you want to make sure the model is actually using the context you are sending it, you are best off curating it first and only sending the most relevant context. In other words, doing traditional RAG.",[20,6448,6449],{},"Here in this template you will use the Gemini 1.5 Flash but you can also use other multimodal models by gemini accordingly.",[6209,6451],{"alt":6452,"src":6453},"Gemini 1.5 flash overview","/assets/content/showcases/gemini_rag/gemini1.5flashtable.png",[64,6455,6457],{"id":6456},"multimodality-with-gemini-15-flash","Multimodality with Gemini-1.5-Flash",[20,6459,6460],{},"Gemini 1.5 Flash is the newest addition to the Gemini family of large language models, and it’s specifically designed to be fast, efficient, and cost-effective for high-volume tasks. This is achieved by being a lighter model than the Gemini 1.5 Pro.",[20,6462,6463],{},"According to the paper from Google DeepMind, Gemini 1.5 Flash is “a more lightweight variant designed for efficiency with minimal regression in quality” and uses the transformer decoder model architecture “and multimodal capabilities as Gemini 1.5 Pro, designed for efficient utilization of tensor processing units (TPUs) with lower latency for model serving.”",[64,6465,6467],{"id":6466},"gemini-15-flash-key-features","Gemini 1.5 Flash: Key Features",[69,6469,6470,6476,6482,6488,6494],{},[72,6471,6472,6475],{},[36,6473,6474],{},"Speed and Efficiency",": Fastest Gemini model at 60 tokens/second, ideal for real-time tasks, reducing costs by delaying autoscaling.",[72,6477,6478,6481],{},[36,6479,6480],{},"Cost-Effective",": 1/10 the price of Gemini 1.5 Pro and cheaper than GPT-3.5.",[72,6483,6484,6487],{},[36,6485,6486],{},"Long Context Window",": Processes up to one million tokens, handling one hour of video, 11 hours of audio, or 700,000 words without losing accuracy.",[72,6489,6490,6493],{},[36,6491,6492],{},"Multimodal Reasoning",": Understands text, images, audio, video, PDFs, and tables. Supports function calling and real-time data access.",[72,6495,6496,6499],{},[36,6497,6498],{},"Great Performance",": High performance with large context windows, excelling in long-document QA, long-video QA, and long-context ASR.",[6209,6501],{"alt":6452,"src":6502},"/assets/content/showcases/gemini_rag/gemini1.5flashdetails.png",[15,6504,6506],{"id":6505},"hands-on-multimodal-rag-with-google-gemini","Hands on Multimodal RAG with Google Gemini",[20,6508,6509],{},[3130,6510],{"alt":6511,"src":6512},"Gemini RAG overview","/assets/content/showcases/gemini_rag/RAG_diagram.png",[64,6514,6516],{"id":6515},"step-1-installation","Step 1: Installation",[20,6518,6519,6520,6523],{},"First, we need to install the required packages: pathway",[392,6521,6522],{},"all",", litellm==1.40.0 and google-generativeai.",[6525,6526,6530],"pre",{"className":6527,"code":6528,"language":6529,"meta":201,"style":201},"language-python shiki shiki-themes material-theme-palenight","!pip install 'pathway[all]>=0.14.0' litellm==1.40.0\n","python",[6150,6531,6532],{"__ignoreMap":201},[392,6533,6536,6540,6544,6548,6550,6553,6556,6560,6562],{"class":6534,"line":6535},"line",1,[392,6537,6539],{"class":6538},"s0W1g","!pip install ",[392,6541,6543],{"class":6542},"sAklC","'",[392,6545,6547],{"class":6546},"sfyAc","pathway[all]>=0.14.0",[392,6549,6543],{"class":6542},[392,6551,6552],{"class":6538}," litellm",[392,6554,6555],{"class":6542},"==",[392,6557,6559],{"class":6558},"sx098","1.40",[392,6561,3267],{"class":6542},[392,6563,6565],{"class":6564},"s-wAU","0\n",[64,6567,6569],{"id":6568},"step-2-imports-and-environment-setup","Step 2: Imports and Environment Setup",[20,6571,6572],{},"Next, we import the necessary libraries and set up the environment variables.",[6525,6574,6576],{"className":6527,"code":6575,"language":6529,"meta":201,"style":201},"import logging\nimport os\n",[6150,6577,6578,6587],{"__ignoreMap":201},[392,6579,6580,6584],{"class":6534,"line":6535},[392,6581,6583],{"class":6582},"s6cf3","import",[392,6585,6586],{"class":6538}," logging\n",[392,6588,6589,6591],{"class":6534,"line":510},[392,6590,6583],{"class":6582},[392,6592,6593],{"class":6538}," os\n",[6525,6595,6597],{"className":6527,"code":6596,"language":6529,"meta":201,"style":201},"import google.generativeai as genai\n",[6150,6598,6599],{"__ignoreMap":201},[392,6600,6601,6603,6606,6608,6611,6614],{"class":6534,"line":6535},[392,6602,6583],{"class":6582},[392,6604,6605],{"class":6538}," google",[392,6607,3267],{"class":6542},[392,6609,6610],{"class":6564},"generativeai",[392,6612,6613],{"class":6582}," as",[392,6615,6616],{"class":6538}," genai\n",[6525,6618,6620],{"className":6527,"code":6619,"language":6529,"meta":201,"style":201},"import litellm\n\nimport pathway as pw\n\nfrom pathway.udfs import DiskCache, ExponentialBackoffRetryStrategy\nfrom pathway.xpacks.llm import embedders, llms, parsers, prompts, splitters\nfrom pathway.xpacks.llm.question_answering import BaseRAGQuestionAnswerer\nfrom pathway.xpacks.llm.vector_store import VectorStoreServer\n\n# Set the logging level for LiteLLM to DEBUG\nos.environ[\"LITELLM_LOG\"] = \"DEBUG\"  # to help in debugging\n",[6150,6621,6622,6629,6634,6647,6652,6677,6719,6744,6769,6774,6781],{"__ignoreMap":201},[392,6623,6624,6626],{"class":6534,"line":6535},[392,6625,6583],{"class":6582},[392,6627,6628],{"class":6538}," litellm\n",[392,6630,6631],{"class":6534,"line":510},[392,6632,6633],{"emptyLinePlaceholder":931},"\n",[392,6635,6636,6638,6641,6644],{"class":6534,"line":897},[392,6637,6583],{"class":6582},[392,6639,6640],{"class":6538}," pathway ",[392,6642,6643],{"class":6582},"as",[392,6645,6646],{"class":6538}," pw\n",[392,6648,6650],{"class":6534,"line":6649},4,[392,6651,6633],{"emptyLinePlaceholder":931},[392,6653,6655,6658,6661,6663,6666,6668,6671,6674],{"class":6534,"line":6654},5,[392,6656,6657],{"class":6582},"from",[392,6659,6660],{"class":6538}," pathway",[392,6662,3267],{"class":6542},[392,6664,6665],{"class":6538},"udfs ",[392,6667,6583],{"class":6582},[392,6669,6670],{"class":6538}," DiskCache",[392,6672,6673],{"class":6542},",",[392,6675,6676],{"class":6538}," ExponentialBackoffRetryStrategy\n",[392,6678,6680,6682,6684,6686,6689,6691,6694,6696,6699,6701,6704,6706,6709,6711,6714,6716],{"class":6534,"line":6679},6,[392,6681,6657],{"class":6582},[392,6683,6660],{"class":6538},[392,6685,3267],{"class":6542},[392,6687,6688],{"class":6538},"xpacks",[392,6690,3267],{"class":6542},[392,6692,6693],{"class":6538},"llm ",[392,6695,6583],{"class":6582},[392,6697,6698],{"class":6538}," embedders",[392,6700,6673],{"class":6542},[392,6702,6703],{"class":6538}," llms",[392,6705,6673],{"class":6542},[392,6707,6708],{"class":6538}," parsers",[392,6710,6673],{"class":6542},[392,6712,6713],{"class":6538}," prompts",[392,6715,6673],{"class":6542},[392,6717,6718],{"class":6538}," splitters\n",[392,6720,6722,6724,6726,6728,6730,6732,6734,6736,6739,6741],{"class":6534,"line":6721},7,[392,6723,6657],{"class":6582},[392,6725,6660],{"class":6538},[392,6727,3267],{"class":6542},[392,6729,6688],{"class":6538},[392,6731,3267],{"class":6542},[392,6733,982],{"class":6538},[392,6735,3267],{"class":6542},[392,6737,6738],{"class":6538},"question_answering ",[392,6740,6583],{"class":6582},[392,6742,6743],{"class":6538}," BaseRAGQuestionAnswerer\n",[392,6745,6747,6749,6751,6753,6755,6757,6759,6761,6764,6766],{"class":6534,"line":6746},8,[392,6748,6657],{"class":6582},[392,6750,6660],{"class":6538},[392,6752,3267],{"class":6542},[392,6754,6688],{"class":6538},[392,6756,3267],{"class":6542},[392,6758,982],{"class":6538},[392,6760,3267],{"class":6542},[392,6762,6763],{"class":6538},"vector_store ",[392,6765,6583],{"class":6582},[392,6767,6768],{"class":6538}," VectorStoreServer\n",[392,6770,6772],{"class":6534,"line":6771},9,[392,6773,6633],{"emptyLinePlaceholder":931},[392,6775,6777],{"class":6534,"line":6776},10,[392,6778,6780],{"class":6779},"saEQR","# Set the logging level for LiteLLM to DEBUG\n",[392,6782,6784,6787,6789,6792,6795,6798,6801,6803,6806,6809,6812,6815,6817],{"class":6534,"line":6783},11,[392,6785,6786],{"class":6538},"os",[392,6788,3267],{"class":6542},[392,6790,6791],{"class":6564},"environ",[392,6793,6794],{"class":6542},"[",[392,6796,6797],{"class":6542},"\"",[392,6799,6800],{"class":6546},"LITELLM_LOG",[392,6802,6797],{"class":6542},[392,6804,6805],{"class":6542},"]",[392,6807,6808],{"class":6542}," =",[392,6810,6811],{"class":6542}," \"",[392,6813,6814],{"class":6546},"DEBUG",[392,6816,6797],{"class":6542},[392,6818,6819],{"class":6779},"  # to help in debugging\n",[64,6821,6823],{"id":6822},"step-3-api-key-setup-and-license-key-setup","Step 3: API Key Setup and License Key Setup",[20,6825,6826],{},"Set up the API key and the Pathway license key:",[6525,6828,6830],{"className":6527,"code":6829,"language":6529,"meta":201,"style":201},"# Api key setup\nGEMINI_API_KEY = \"Paste your Gemini API Key here\"\n\nos.environ[\"GEMINI_API_KEY\"] = GEMINI_API_KEY\nos.environ[\"TESSDATA_PREFIX\"] = \"/usr/share/tesseract/tessdata/\"\ngenai.configure(api_key=GEMINI_API_KEY)\n\n# License key setup\npw.set_license_key(\"demo-license-key-with-telemetry\")\n\nlogging.basicConfig(\n    level=logging.INFO, format=\"%(asctime)s - %(levelname)s - %(message)s\"\n)\n",[6150,6831,6832,6837,6853,6857,6881,6909,6934,6938,6943,6964,6968,6981,7020],{"__ignoreMap":201},[392,6833,6834],{"class":6534,"line":6535},[392,6835,6836],{"class":6779},"# Api key setup\n",[392,6838,6839,6842,6845,6847,6850],{"class":6534,"line":510},[392,6840,6841],{"class":6538},"GEMINI_API_KEY ",[392,6843,6844],{"class":6542},"=",[392,6846,6811],{"class":6542},[392,6848,6849],{"class":6546},"Paste your Gemini API Key here",[392,6851,6852],{"class":6542},"\"\n",[392,6854,6855],{"class":6534,"line":897},[392,6856,6633],{"emptyLinePlaceholder":931},[392,6858,6859,6861,6863,6865,6867,6869,6872,6874,6876,6878],{"class":6534,"line":6649},[392,6860,6786],{"class":6538},[392,6862,3267],{"class":6542},[392,6864,6791],{"class":6564},[392,6866,6794],{"class":6542},[392,6868,6797],{"class":6542},[392,6870,6871],{"class":6546},"GEMINI_API_KEY",[392,6873,6797],{"class":6542},[392,6875,6805],{"class":6542},[392,6877,6808],{"class":6542},[392,6879,6880],{"class":6538}," GEMINI_API_KEY\n",[392,6882,6883,6885,6887,6889,6891,6893,6896,6898,6900,6902,6904,6907],{"class":6534,"line":6654},[392,6884,6786],{"class":6538},[392,6886,3267],{"class":6542},[392,6888,6791],{"class":6564},[392,6890,6794],{"class":6542},[392,6892,6797],{"class":6542},[392,6894,6895],{"class":6546},"TESSDATA_PREFIX",[392,6897,6797],{"class":6542},[392,6899,6805],{"class":6542},[392,6901,6808],{"class":6542},[392,6903,6811],{"class":6542},[392,6905,6906],{"class":6546},"/usr/share/tesseract/tessdata/",[392,6908,6852],{"class":6542},[392,6910,6911,6914,6916,6920,6923,6927,6929,6931],{"class":6534,"line":6679},[392,6912,6913],{"class":6538},"genai",[392,6915,3267],{"class":6542},[392,6917,6919],{"class":6918},"sdLwU","configure",[392,6921,6922],{"class":6542},"(",[392,6924,6926],{"class":6925},"s7ZW3","api_key",[392,6928,6844],{"class":6542},[392,6930,6871],{"class":6918},[392,6932,6933],{"class":6542},")\n",[392,6935,6936],{"class":6534,"line":6721},[392,6937,6633],{"emptyLinePlaceholder":931},[392,6939,6940],{"class":6534,"line":6746},[392,6941,6942],{"class":6779},"# License key setup\n",[392,6944,6945,6948,6950,6953,6955,6957,6960,6962],{"class":6534,"line":6771},[392,6946,6947],{"class":6538},"pw",[392,6949,3267],{"class":6542},[392,6951,6952],{"class":6918},"set_license_key",[392,6954,6922],{"class":6542},[392,6956,6797],{"class":6542},[392,6958,6959],{"class":6546},"demo-license-key-with-telemetry",[392,6961,6797],{"class":6542},[392,6963,6933],{"class":6542},[392,6965,6966],{"class":6534,"line":6776},[392,6967,6633],{"emptyLinePlaceholder":931},[392,6969,6970,6973,6975,6978],{"class":6534,"line":6783},[392,6971,6972],{"class":6538},"logging",[392,6974,3267],{"class":6542},[392,6976,6977],{"class":6918},"basicConfig",[392,6979,6980],{"class":6542},"(\n",[392,6982,6984,6987,6989,6991,6993,6996,6998,7001,7003,7005,7008,7010,7013,7015,7018],{"class":6534,"line":6983},12,[392,6985,6986],{"class":6925},"    level",[392,6988,6844],{"class":6542},[392,6990,6972],{"class":6918},[392,6992,3267],{"class":6542},[392,6994,6995],{"class":6564},"INFO",[392,6997,6673],{"class":6542},[392,6999,7000],{"class":6925}," format",[392,7002,6844],{"class":6542},[392,7004,6797],{"class":6542},[392,7006,7007],{"class":6558},"%(asctime)s",[392,7009,5794],{"class":6546},[392,7011,7012],{"class":6558},"%(levelname)s",[392,7014,5794],{"class":6546},[392,7016,7017],{"class":6558},"%(message)s",[392,7019,6852],{"class":6542},[392,7021,7023],{"class":6534,"line":7022},13,[392,7024,6933],{"class":6542},[64,7026,7028],{"id":7027},"step-4-upload-your-file","Step 4: Upload your file",[20,7030,7031,7032,7035],{},"Create a ",[6150,7033,7034],{},"./data"," directory if it doesn't already exist. This is where the uploaded files will be stored. Then upload your pdf documents.",[20,7037,7038,7039,7042],{},"You can also omit this cell if you are running locally on your system - in that case create a ",[6150,7040,7041],{},"data"," folder in the current directory and copy the files and comment out this cell.",[6525,7044,7046],{"className":6527,"code":7045,"language":6529,"meta":201,"style":201},"!mkdir -p data\n",[6150,7047,7048],{"__ignoreMap":201},[392,7049,7050,7053,7056],{"class":6534,"line":6535},[392,7051,7052],{"class":6538},"!mkdir ",[392,7054,7055],{"class":6542},"-",[392,7057,7058],{"class":6538},"p data\n",[6525,7060,7062],{"className":6527,"code":7061,"language":6529,"meta":201,"style":201},"# Demo pdf for testing\n!wget -q -P ./data/ https://github.com/pathwaycom/llm-app/raw/main/templates/multimodal_rag/data/20230203_alphabet_10K.pdf\n",[6150,7063,7064,7069],{"__ignoreMap":201},[392,7065,7066],{"class":6534,"line":6535},[392,7067,7068],{"class":6779},"# Demo pdf for testing\n",[392,7070,7071,7074,7076,7079,7081,7084,7087,7089,7091,7094,7097,7100,7102,7105,7107,7110,7112,7114,7116,7119,7121,7124,7126,7129,7131,7134,7136,7139,7141,7143,7145,7148,7150],{"class":6534,"line":510},[392,7072,7073],{"class":6538},"!wget ",[392,7075,7055],{"class":6542},[392,7077,7078],{"class":6538},"q ",[392,7080,7055],{"class":6542},[392,7082,7083],{"class":6538},"P ",[392,7085,7086],{"class":6542},"./",[392,7088,7041],{"class":6538},[392,7090,3284],{"class":6542},[392,7092,7093],{"class":6538}," https",[392,7095,7096],{"class":6542},"://",[392,7098,7099],{"class":6538},"github",[392,7101,3267],{"class":6542},[392,7103,7104],{"class":6564},"com",[392,7106,3284],{"class":6542},[392,7108,7109],{"class":6538},"pathwaycom",[392,7111,3284],{"class":6542},[392,7113,982],{"class":6538},[392,7115,7055],{"class":6542},[392,7117,7118],{"class":6538},"app",[392,7120,3284],{"class":6542},[392,7122,7123],{"class":6538},"raw",[392,7125,3284],{"class":6542},[392,7127,7128],{"class":6538},"main",[392,7130,3284],{"class":6542},[392,7132,7133],{"class":6538},"templates",[392,7135,3284],{"class":6542},[392,7137,7138],{"class":6538},"multimodal_rag",[392,7140,3284],{"class":6542},[392,7142,7041],{"class":6538},[392,7144,3284],{"class":6542},[392,7146,7147],{"class":6538},"20230203_alphabet_10K",[392,7149,3267],{"class":6542},[392,7151,7152],{"class":6564},"pdf\n",[6314,7154,7156],{"id":7155},"reading-pdf-data","Reading PDF Data",[20,7158,7159],{},"Next, we read the PDF data from a folder.",[6525,7161,7163],{"className":6527,"code":7162,"language":6529,"meta":201,"style":201},"# Read the PDF data\nfolder = pw.io.fs.read(\n    path=\"./data/\",\n    format=\"binary\",\n    with_metadata=True,\n)\nsources = [folder]  # you can add any other Pathway connector here!\n",[6150,7164,7165,7170,7197,7214,7230,7238,7242],{"__ignoreMap":201},[392,7166,7167],{"class":6534,"line":6535},[392,7168,7169],{"class":6779},"# Read the PDF data\n",[392,7171,7172,7175,7177,7180,7182,7185,7187,7190,7192,7195],{"class":6534,"line":510},[392,7173,7174],{"class":6538},"folder ",[392,7176,6844],{"class":6542},[392,7178,7179],{"class":6538}," pw",[392,7181,3267],{"class":6542},[392,7183,7184],{"class":6564},"io",[392,7186,3267],{"class":6542},[392,7188,7189],{"class":6564},"fs",[392,7191,3267],{"class":6542},[392,7193,7194],{"class":6918},"read",[392,7196,6980],{"class":6542},[392,7198,7199,7202,7204,7206,7209,7211],{"class":6534,"line":897},[392,7200,7201],{"class":6925},"    path",[392,7203,6844],{"class":6542},[392,7205,6797],{"class":6542},[392,7207,7208],{"class":6546},"./data/",[392,7210,6797],{"class":6542},[392,7212,7213],{"class":6542},",\n",[392,7215,7216,7219,7221,7223,7226,7228],{"class":6534,"line":6649},[392,7217,7218],{"class":6925},"    format",[392,7220,6844],{"class":6542},[392,7222,6797],{"class":6542},[392,7224,7225],{"class":6546},"binary",[392,7227,6797],{"class":6542},[392,7229,7213],{"class":6542},[392,7231,7232,7235],{"class":6534,"line":6654},[392,7233,7234],{"class":6925},"    with_metadata",[392,7236,7237],{"class":6542},"=True,\n",[392,7239,7240],{"class":6534,"line":6679},[392,7241,6933],{"class":6542},[392,7243,7244,7247,7249,7252,7255,7257],{"class":6534,"line":6721},[392,7245,7246],{"class":6538},"sources ",[392,7248,6844],{"class":6542},[392,7250,7251],{"class":6542}," [",[392,7253,7254],{"class":6538},"folder",[392,7256,6805],{"class":6542},[392,7258,7259],{"class":6779},"  # you can add any other Pathway connector here!\n",[64,7261,7263],{"id":7262},"step-5-document-processing-and-question-answering-setup","Step 5: Document Processing and Question Answering Setup",[6314,7265,7267],{"id":7266},"setting-up-litellm-chat","Setting Up LiteLLM Chat",[20,7269,7270],{},"Set up a LiteLLM chat instance with retry and cache strategies:",[6525,7272,7274],{"className":6527,"code":7273,"language":6529,"meta":201,"style":201},"# Setup LiteLLM chat\nchat = llms.LiteLLMChat(\n    model=\"gemini/gemini-1.5-flash\",  # Model specified for LiteLLM\n    retry_strategy=ExponentialBackoffRetryStrategy(max_retries=6, backoff_factor=2.5),\n    temperature=0.0,\n)\n",[6150,7275,7276,7281,7297,7316,7349,7361],{"__ignoreMap":201},[392,7277,7278],{"class":6534,"line":6535},[392,7279,7280],{"class":6779},"# Setup LiteLLM chat\n",[392,7282,7283,7286,7288,7290,7292,7295],{"class":6534,"line":510},[392,7284,7285],{"class":6538},"chat ",[392,7287,6844],{"class":6542},[392,7289,6703],{"class":6538},[392,7291,3267],{"class":6542},[392,7293,7294],{"class":6918},"LiteLLMChat",[392,7296,6980],{"class":6542},[392,7298,7299,7302,7304,7306,7309,7311,7313],{"class":6534,"line":897},[392,7300,7301],{"class":6925},"    model",[392,7303,6844],{"class":6542},[392,7305,6797],{"class":6542},[392,7307,7308],{"class":6546},"gemini/gemini-1.5-flash",[392,7310,6797],{"class":6542},[392,7312,6673],{"class":6542},[392,7314,7315],{"class":6779},"  # Model specified for LiteLLM\n",[392,7317,7318,7321,7323,7326,7328,7331,7333,7336,7338,7341,7343,7346],{"class":6534,"line":6649},[392,7319,7320],{"class":6925},"    retry_strategy",[392,7322,6844],{"class":6542},[392,7324,7325],{"class":6918},"ExponentialBackoffRetryStrategy",[392,7327,6922],{"class":6542},[392,7329,7330],{"class":6925},"max_retries",[392,7332,6844],{"class":6542},[392,7334,7335],{"class":6558},"6",[392,7337,6673],{"class":6542},[392,7339,7340],{"class":6925}," backoff_factor",[392,7342,6844],{"class":6542},[392,7344,7345],{"class":6558},"2.5",[392,7347,7348],{"class":6542},"),\n",[392,7350,7351,7354,7356,7359],{"class":6534,"line":6654},[392,7352,7353],{"class":6925},"    temperature",[392,7355,6844],{"class":6542},[392,7357,7358],{"class":6558},"0.0",[392,7360,7213],{"class":6542},[392,7362,7363],{"class":6534,"line":6679},[392,7364,6933],{"class":6542},[6314,7366,7368],{"id":7367},"setting-up-embedder","Setting Up Embedder",[20,7370,7371,7372,7375,7376,7379,7380,7383],{},"Let's utilize Gemini embedders. The ",[6150,7373,7374],{},"GeminiEmbedder"," class in Pathway provides an interface for interacting with Gemini embedders. It generates semantic embeddings with a specified model, providing methods for single items (",[6150,7377,7378],{},"embed","), batches (",[6150,7381,7382],{},"embed_batch","), and direct calls.",[6525,7385,7387],{"className":6527,"code":7386,"language":6529,"meta":201,"style":201},"# Setup embedder\nembedder = embedders.GeminiEmbedder(\n    model=\"models/embedding-001\",\n    retry_strategy=ExponentialBackoffRetryStrategy(max_retries=6, backoff_factor=2.5),\n)  # Specify embedder here\n",[6150,7388,7389,7394,7409,7424,7450],{"__ignoreMap":201},[392,7390,7391],{"class":6534,"line":6535},[392,7392,7393],{"class":6779},"# Setup embedder\n",[392,7395,7396,7399,7401,7403,7405,7407],{"class":6534,"line":510},[392,7397,7398],{"class":6538},"embedder ",[392,7400,6844],{"class":6542},[392,7402,6698],{"class":6538},[392,7404,3267],{"class":6542},[392,7406,7374],{"class":6918},[392,7408,6980],{"class":6542},[392,7410,7411,7413,7415,7417,7420,7422],{"class":6534,"line":897},[392,7412,7301],{"class":6925},[392,7414,6844],{"class":6542},[392,7416,6797],{"class":6542},[392,7418,7419],{"class":6546},"models/embedding-001",[392,7421,6797],{"class":6542},[392,7423,7213],{"class":6542},[392,7425,7426,7428,7430,7432,7434,7436,7438,7440,7442,7444,7446,7448],{"class":6534,"line":6649},[392,7427,7320],{"class":6925},[392,7429,6844],{"class":6542},[392,7431,7325],{"class":6918},[392,7433,6922],{"class":6542},[392,7435,7330],{"class":6925},[392,7437,6844],{"class":6542},[392,7439,7335],{"class":6558},[392,7441,6673],{"class":6542},[392,7443,7340],{"class":6925},[392,7445,6844],{"class":6542},[392,7447,7345],{"class":6558},[392,7449,7348],{"class":6542},[392,7451,7452,7454],{"class":6534,"line":6654},[392,7453,3286],{"class":6542},[392,7455,7456],{"class":6779},"  # Specify embedder here\n",[6314,7458,7460],{"id":7459},"setting-up-parser","Setting Up Parser",[20,7462,7463],{},"Next, we set up a parser for the document store.",[6525,7465,7467],{"className":6527,"code":7466,"language":6529,"meta":201,"style":201},"# Setup parser\ntable_args = {\n    \"parsing_algorithm\": \"llm\",  # for tables\n    \"llm\": chat,\n    \"prompt\": prompts.DEFAULT_MD_TABLE_PARSE_PROMPT,\n}\n\nimage_args = {\n    \"parsing_algorithm\": \"llm\",  # for images\n    \"llm\": chat,\n    \"prompt\": prompts.DEFAULT_IMAGE_PARSE_PROMPT,\n}\n\nparser = parsers.DoclingParser(multimodal_llm=chat)\n",[6150,7468,7469,7474,7484,7508,7523,7543,7548,7552,7561,7582,7596,7615,7619,7623],{"__ignoreMap":201},[392,7470,7471],{"class":6534,"line":6535},[392,7472,7473],{"class":6779},"# Setup parser\n",[392,7475,7476,7479,7481],{"class":6534,"line":510},[392,7477,7478],{"class":6538},"table_args ",[392,7480,6844],{"class":6542},[392,7482,7483],{"class":6542}," {\n",[392,7485,7486,7489,7492,7494,7497,7499,7501,7503,7505],{"class":6534,"line":897},[392,7487,7488],{"class":6542},"    \"",[392,7490,7491],{"class":6546},"parsing_algorithm",[392,7493,6797],{"class":6542},[392,7495,7496],{"class":6542},":",[392,7498,6811],{"class":6542},[392,7500,982],{"class":6546},[392,7502,6797],{"class":6542},[392,7504,6673],{"class":6542},[392,7506,7507],{"class":6779},"  # for tables\n",[392,7509,7510,7512,7514,7516,7518,7521],{"class":6534,"line":6649},[392,7511,7488],{"class":6542},[392,7513,982],{"class":6546},[392,7515,6797],{"class":6542},[392,7517,7496],{"class":6542},[392,7519,7520],{"class":6538}," chat",[392,7522,7213],{"class":6542},[392,7524,7525,7527,7530,7532,7534,7536,7538,7541],{"class":6534,"line":6654},[392,7526,7488],{"class":6542},[392,7528,7529],{"class":6546},"prompt",[392,7531,6797],{"class":6542},[392,7533,7496],{"class":6542},[392,7535,6713],{"class":6538},[392,7537,3267],{"class":6542},[392,7539,7540],{"class":6564},"DEFAULT_MD_TABLE_PARSE_PROMPT",[392,7542,7213],{"class":6542},[392,7544,7545],{"class":6534,"line":6679},[392,7546,7547],{"class":6542},"}\n",[392,7549,7550],{"class":6534,"line":6721},[392,7551,6633],{"emptyLinePlaceholder":931},[392,7553,7554,7557,7559],{"class":6534,"line":6746},[392,7555,7556],{"class":6538},"image_args ",[392,7558,6844],{"class":6542},[392,7560,7483],{"class":6542},[392,7562,7563,7565,7567,7569,7571,7573,7575,7577,7579],{"class":6534,"line":6771},[392,7564,7488],{"class":6542},[392,7566,7491],{"class":6546},[392,7568,6797],{"class":6542},[392,7570,7496],{"class":6542},[392,7572,6811],{"class":6542},[392,7574,982],{"class":6546},[392,7576,6797],{"class":6542},[392,7578,6673],{"class":6542},[392,7580,7581],{"class":6779},"  # for images\n",[392,7583,7584,7586,7588,7590,7592,7594],{"class":6534,"line":6776},[392,7585,7488],{"class":6542},[392,7587,982],{"class":6546},[392,7589,6797],{"class":6542},[392,7591,7496],{"class":6542},[392,7593,7520],{"class":6538},[392,7595,7213],{"class":6542},[392,7597,7598,7600,7602,7604,7606,7608,7610,7613],{"class":6534,"line":6783},[392,7599,7488],{"class":6542},[392,7601,7529],{"class":6546},[392,7603,6797],{"class":6542},[392,7605,7496],{"class":6542},[392,7607,6713],{"class":6538},[392,7609,3267],{"class":6542},[392,7611,7612],{"class":6564},"DEFAULT_IMAGE_PARSE_PROMPT",[392,7614,7213],{"class":6542},[392,7616,7617],{"class":6534,"line":6983},[392,7618,7547],{"class":6542},[392,7620,7621],{"class":6534,"line":7022},[392,7622,6633],{"emptyLinePlaceholder":931},[392,7624,7626,7629,7631,7633,7635,7638,7640,7643,7645,7648],{"class":6534,"line":7625},14,[392,7627,7628],{"class":6538},"parser ",[392,7630,6844],{"class":6542},[392,7632,6708],{"class":6538},[392,7634,3267],{"class":6542},[392,7636,7637],{"class":6918},"DoclingParser",[392,7639,6922],{"class":6542},[392,7641,7642],{"class":6925},"multimodal_llm",[392,7644,6844],{"class":6542},[392,7646,7647],{"class":6918},"chat",[392,7649,6933],{"class":6542},[6314,7651,7653],{"id":7652},"setting-up-document-store","Setting Up Document Store",[20,7655,7656],{},"We will set up the document store with the sources, embedder, and parser.",[6525,7658,7660],{"className":6527,"code":7659,"language":6529,"meta":201,"style":201},"# Setup document store\n# splitter = splitters.TokenCountSplitter()\ndoc_store = VectorStoreServer(\n    *sources,\n    embedder=embedder,\n    splitter=splitter,\n    parser=parser,\n)\n",[6150,7661,7662,7667,7672,7684,7694,7706,7718,7730],{"__ignoreMap":201},[392,7663,7664],{"class":6534,"line":6535},[392,7665,7666],{"class":6779},"# Setup document store\n",[392,7668,7669],{"class":6534,"line":510},[392,7670,7671],{"class":6779},"# splitter = splitters.TokenCountSplitter()\n",[392,7673,7674,7677,7679,7682],{"class":6534,"line":897},[392,7675,7676],{"class":6538},"doc_store ",[392,7678,6844],{"class":6542},[392,7680,7681],{"class":6918}," VectorStoreServer",[392,7683,6980],{"class":6542},[392,7685,7686,7689,7692],{"class":6534,"line":6649},[392,7687,7688],{"class":6542},"    *",[392,7690,7691],{"class":6918},"sources",[392,7693,7213],{"class":6542},[392,7695,7696,7699,7701,7704],{"class":6534,"line":6654},[392,7697,7698],{"class":6925},"    embedder",[392,7700,6844],{"class":6542},[392,7702,7703],{"class":6918},"embedder",[392,7705,7213],{"class":6542},[392,7707,7708,7711,7713,7716],{"class":6534,"line":6679},[392,7709,7710],{"class":6925},"    splitter",[392,7712,6844],{"class":6542},[392,7714,7715],{"class":6918},"splitter",[392,7717,7213],{"class":6542},[392,7719,7720,7723,7725,7728],{"class":6534,"line":6721},[392,7721,7722],{"class":6925},"    parser",[392,7724,6844],{"class":6542},[392,7726,7727],{"class":6918},"parser",[392,7729,7213],{"class":6542},[392,7731,7732],{"class":6534,"line":6746},[392,7733,6933],{"class":6542},[64,7735,7737],{"id":7736},"step-6-setting-up-question-answerer-application","Step 6: Setting Up Question Answerer Application",[20,7739,7740],{},"We will set up the question answerer application using the LiteLLM-based chat object.",[6525,7742,7744],{"className":6527,"code":7743,"language":6529,"meta":201,"style":201},"# Setup question answerer application\napp = BaseRAGQuestionAnswerer(\n        llm=chat,  # Using the LiteLLM-based chat object\n        indexer=doc_store, search_topk=2,\n        short_prompt_template=prompts.prompt_qa)\n",[6150,7745,7746,7751,7763,7777,7798],{"__ignoreMap":201},[392,7747,7748],{"class":6534,"line":6535},[392,7749,7750],{"class":6779},"# Setup question answerer application\n",[392,7752,7753,7756,7758,7761],{"class":6534,"line":510},[392,7754,7755],{"class":6538},"app ",[392,7757,6844],{"class":6542},[392,7759,7760],{"class":6918}," BaseRAGQuestionAnswerer",[392,7762,6980],{"class":6542},[392,7764,7765,7768,7770,7772,7774],{"class":6534,"line":897},[392,7766,7767],{"class":6925},"        llm",[392,7769,6844],{"class":6542},[392,7771,7647],{"class":6918},[392,7773,6673],{"class":6542},[392,7775,7776],{"class":6779},"  # Using the LiteLLM-based chat object\n",[392,7778,7779,7782,7784,7787,7789,7792,7794,7796],{"class":6534,"line":6649},[392,7780,7781],{"class":6925},"        indexer",[392,7783,6844],{"class":6542},[392,7785,7786],{"class":6918},"doc_store",[392,7788,6673],{"class":6542},[392,7790,7791],{"class":6925}," search_topk",[392,7793,6844],{"class":6542},[392,7795,514],{"class":6558},[392,7797,7213],{"class":6542},[392,7799,7800,7803,7805,7808,7810,7813],{"class":6534,"line":6654},[392,7801,7802],{"class":6925},"        short_prompt_template",[392,7804,6844],{"class":6542},[392,7806,7807],{"class":6918},"prompts",[392,7809,3267],{"class":6542},[392,7811,7812],{"class":6564},"prompt_qa",[392,7814,6933],{"class":6542},[6314,7816,7818],{"id":7817},"building-and-running-the-server","Building and Running the Server",[20,7820,7821],{},"Finally, we build and run the server.",[6525,7823,7825],{"className":6527,"code":7824,"language":6529,"meta":201,"style":201},"# Build and run the server\napp_host = \"0.0.0.0\"\napp_port = 8000\napp.build_server(host=app_host, port=app_port)\n",[6150,7826,7827,7832,7846,7856],{"__ignoreMap":201},[392,7828,7829],{"class":6534,"line":6535},[392,7830,7831],{"class":6779},"# Build and run the server\n",[392,7833,7834,7837,7839,7841,7844],{"class":6534,"line":510},[392,7835,7836],{"class":6538},"app_host ",[392,7838,6844],{"class":6542},[392,7840,6811],{"class":6542},[392,7842,7843],{"class":6546},"0.0.0.0",[392,7845,6852],{"class":6542},[392,7847,7848,7851,7853],{"class":6534,"line":897},[392,7849,7850],{"class":6538},"app_port ",[392,7852,6844],{"class":6542},[392,7854,7855],{"class":6558}," 8000\n",[392,7857,7858,7860,7862,7865,7867,7870,7872,7875,7877,7880,7882,7885],{"class":6534,"line":6649},[392,7859,7118],{"class":6538},[392,7861,3267],{"class":6542},[392,7863,7864],{"class":6918},"build_server",[392,7866,6922],{"class":6542},[392,7868,7869],{"class":6925},"host",[392,7871,6844],{"class":6542},[392,7873,7874],{"class":6918},"app_host",[392,7876,6673],{"class":6542},[392,7878,7879],{"class":6925}," port",[392,7881,6844],{"class":6542},[392,7883,7884],{"class":6918},"app_port",[392,7886,6933],{"class":6542},[6525,7888,7890],{"className":6527,"code":7889,"language":6529,"meta":201,"style":201},"import threading\nt = threading.Thread(target=app.run_server, name=\"BaseRAGQuestionAnswerer\")\nt.daemon = True\nthr = t.start()\n",[6150,7891,7892,7899,7944,7959],{"__ignoreMap":201},[392,7893,7894,7896],{"class":6534,"line":6535},[392,7895,6583],{"class":6582},[392,7897,7898],{"class":6538}," threading\n",[392,7900,7901,7904,7906,7909,7911,7914,7916,7919,7921,7923,7925,7928,7930,7933,7935,7937,7940,7942],{"class":6534,"line":510},[392,7902,7903],{"class":6538},"t ",[392,7905,6844],{"class":6542},[392,7907,7908],{"class":6538}," threading",[392,7910,3267],{"class":6542},[392,7912,7913],{"class":6918},"Thread",[392,7915,6922],{"class":6542},[392,7917,7918],{"class":6925},"target",[392,7920,6844],{"class":6542},[392,7922,7118],{"class":6918},[392,7924,3267],{"class":6542},[392,7926,7927],{"class":6564},"run_server",[392,7929,6673],{"class":6542},[392,7931,7932],{"class":6925}," name",[392,7934,6844],{"class":6542},[392,7936,6797],{"class":6542},[392,7938,7939],{"class":6546},"BaseRAGQuestionAnswerer",[392,7941,6797],{"class":6542},[392,7943,6933],{"class":6542},[392,7945,7946,7949,7951,7954,7956],{"class":6534,"line":897},[392,7947,7948],{"class":6538},"t",[392,7950,3267],{"class":6542},[392,7952,7953],{"class":6564},"daemon",[392,7955,6808],{"class":6542},[392,7957,7958],{"class":6542}," True\n",[392,7960,7961,7964,7966,7969,7971,7974],{"class":6534,"line":6649},[392,7962,7963],{"class":6538},"thr ",[392,7965,6844],{"class":6542},[392,7967,7968],{"class":6538}," t",[392,7970,3267],{"class":6542},[392,7972,7973],{"class":6918},"start",[392,7975,7976],{"class":6542},"()\n",[6525,7978,7980],{"className":6527,"code":7979,"language":6529,"meta":201,"style":201},"from pathway.xpacks.llm.question_answering import RAGClient\n\n# Initialize the RAG client\nclient = RAGClient(host=\"0.0.0.0\", port=8000)\n",[6150,7981,7982,8005,8009,8014],{"__ignoreMap":201},[392,7983,7984,7986,7988,7990,7992,7994,7996,7998,8000,8002],{"class":6534,"line":6535},[392,7985,6657],{"class":6582},[392,7987,6660],{"class":6538},[392,7989,3267],{"class":6542},[392,7991,6688],{"class":6538},[392,7993,3267],{"class":6542},[392,7995,982],{"class":6538},[392,7997,3267],{"class":6542},[392,7999,6738],{"class":6538},[392,8001,6583],{"class":6582},[392,8003,8004],{"class":6538}," RAGClient\n",[392,8006,8007],{"class":6534,"line":510},[392,8008,6633],{"emptyLinePlaceholder":931},[392,8010,8011],{"class":6534,"line":897},[392,8012,8013],{"class":6779},"# Initialize the RAG client\n",[392,8015,8016,8019,8021,8024,8026,8028,8030,8032,8034,8036,8038,8040,8042,8045],{"class":6534,"line":6649},[392,8017,8018],{"class":6538},"client ",[392,8020,6844],{"class":6542},[392,8022,8023],{"class":6918}," RAGClient",[392,8025,6922],{"class":6542},[392,8027,7869],{"class":6925},[392,8029,6844],{"class":6542},[392,8031,6797],{"class":6542},[392,8033,7843],{"class":6546},[392,8035,6797],{"class":6542},[392,8037,6673],{"class":6542},[392,8039,7879],{"class":6925},[392,8041,6844],{"class":6542},[392,8043,8044],{"class":6558},"8000",[392,8046,6933],{"class":6542},[6525,8048,8050],{"className":6527,"code":8049,"language":6529,"meta":201,"style":201},"# Example usage\n\nresponse = client.answer(\"What is the Total Stockholders' equity as of December 31, 2022?\")\nprint(response)\n\n",[6150,8051,8052,8057,8061,8087],{"__ignoreMap":201},[392,8053,8054],{"class":6534,"line":6535},[392,8055,8056],{"class":6779},"# Example usage\n",[392,8058,8059],{"class":6534,"line":510},[392,8060,6633],{"emptyLinePlaceholder":931},[392,8062,8063,8066,8068,8071,8073,8076,8078,8080,8083,8085],{"class":6534,"line":897},[392,8064,8065],{"class":6538},"response ",[392,8067,6844],{"class":6542},[392,8069,8070],{"class":6538}," client",[392,8072,3267],{"class":6542},[392,8074,8075],{"class":6918},"answer",[392,8077,6922],{"class":6542},[392,8079,6797],{"class":6542},[392,8081,8082],{"class":6546},"What is the Total Stockholders' equity as of December 31, 2022?",[392,8084,6797],{"class":6542},[392,8086,6933],{"class":6542},[392,8088,8089,8092,8094,8097],{"class":6534,"line":6649},[392,8090,8091],{"class":6918},"print",[392,8093,6922],{"class":6542},[392,8095,8096],{"class":6918},"response",[392,8098,6933],{"class":6542},[6525,8100,8105],{"className":8101,"code":8103,"language":8104},[8102],"language-text","$256,144 million\n","text",[6150,8106,8103],{"__ignoreMap":201},[20,8108,8109],{},"Now your chatbot is now running live! You can ask any questions and get information from your documents instantly.",[15,8111,705],{"id":704},[20,8113,8114],{},"This article demonstrated how to implement a Multimodal RAG service using Pathway and Gemini. The setup leverages the capabilities of LiteLLM to process and query multimodal data effectively. If you're looking for a cost-effective alternative, consider using the Gemini Mini, which provides great performance at a lower cost.",[20,8116,8117,8118,8120],{},"For more detailed insights and an alternative approach, check out our article on multimodal RAG using GPT-4o ",[717,8119,6256],{"href":6255},". This will give you another perspective on how to handle multimodal RAG applications using different models and techniques.\nBy following the steps outlined above, you can efficiently integrate and utilize various data types to enhance your AI applications, ensuring more accurate and contextually rich outputs.",[8122,8123,8124],"style",{},"html pre.shiki code .s0W1g, html code.shiki .s0W1g{--shiki-default:#BABED8}html pre.shiki code .sAklC, html code.shiki .sAklC{--shiki-default:#89DDFF}html pre.shiki code .sfyAc, html code.shiki .sfyAc{--shiki-default:#C3E88D}html pre.shiki code .sx098, html code.shiki .sx098{--shiki-default:#F78C6C}html pre.shiki code .s-wAU, html code.shiki .s-wAU{--shiki-default:#F07178}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html pre.shiki code .s6cf3, html code.shiki .s6cf3{--shiki-default:#89DDFF;--shiki-default-font-style:italic}html pre.shiki code .saEQR, html code.shiki .saEQR{--shiki-default:#676E95;--shiki-default-font-style:italic}html pre.shiki code .sdLwU, html code.shiki .sdLwU{--shiki-default:#82AAFF}html pre.shiki code .s7ZW3, html code.shiki .s7ZW3{--shiki-default:#BABED8;--shiki-default-font-style:italic}",{"title":201,"searchDepth":510,"depth":510,"links":8126},[8127,8128,8135,8143],{"id":6265,"depth":510,"text":6266},{"id":6295,"depth":510,"text":6296,"children":8129},[8130,8131,8132,8133,8134],{"id":6299,"depth":897,"text":6300},{"id":6370,"depth":897,"text":6283},{"id":6423,"depth":897,"text":6286},{"id":6456,"depth":897,"text":6457},{"id":6466,"depth":897,"text":6467},{"id":6505,"depth":510,"text":6506,"children":8136},[8137,8138,8139,8140,8141,8142],{"id":6515,"depth":897,"text":6516},{"id":6568,"depth":897,"text":6569},{"id":6822,"depth":897,"text":6823},{"id":7027,"depth":897,"text":7028},{"id":7262,"depth":897,"text":7263},{"id":7736,"depth":897,"text":7737},{"id":704,"depth":510,"text":705},"End-to-end template showing how you can launch a document processing RAG pipeline that utilizes Gemini and Pathway",{"aside":931,"layout":925,"thumbnail":8146,"date":979,"tags":8147,"keywords":8148,"notebook_export_path":995,"run_template":996,"hidden":931},{"src":977,"fit":978},[981,982],[984,985,986,987,988,989,990,991,992,993,994],{"title":972,"description":8144},{"loc":973},"pgWjERJ2NvHTeunGcu3MKJlhcra6lra69ndyvdPs00M",{"id":8153,"title":998,"author":8154,"body":8160,"description":201,"extension":923,"meta":9346,"navigation":931,"path":999,"seo":9350,"sitemap":9351,"stem":1000,"__hash__":9352},"content/framework/blog/1002.langchain-integration.md",{"id":8155,"url":8156,"name":8157,"description":8158,"img":1141,"provider":10,"linkedin":8159},"szymon","szymon-dudycz","Szymon Dudycz","Algorithm and Data Processing Magician","https://www.linkedin.com/in/szymon-dudycz-19ab2962/",{"type":12,"value":8161,"toc":9338},[8162,8165,8168,8182,8185,8189,8204,8224,8228,8245,8252,8274,8277,8385,8391,8538,8548,8731,8747,8751,8760,8797,8850,8901,8904,8942,8945,8978,8982,8985,9237,9240,9265,9269,9281,9300,9335],[2612,8163,998],{"id":8164},"langchain-and-pathway-rag-apps-with-always-up-to-date-knowledge",[20,8166,8167],{},"You can now use Pathway in your RAG applications which enables always up-to-date knowledge from your documents to LLMs with Langchaing integration.",[20,8169,8170,8171,8176,8177,3267],{},"Pathway is now available on ",[717,8172,8175],{"href":8173,"rel":8174},"https://python.langchain.com/docs/integrations/vectorstores/pathway/",[733],"Langchain",", a framework for developing applications powered by large language models (LLMs).\nYou can now query Pathway and access up-to-date documents for your RAG applications from LangChain using ",[717,8178,8181],{"href":8179,"rel":8180},"https://api.python.langchain.com/en/latest/vectorstores/langchain_community.vectorstores.pathway.PathwayVectorClient.html",[733],"PathwayVectorClient",[20,8183,8184],{},"With this new integration, you will be able to use Pathway Vector Store natively in LangChain. In this guide, you will have a quick dive into Pathway + LangChain to learn how to create a simple, yet powerful RAG solution.",[15,8186,8188],{"id":8187},"prerequisites","Prerequisites",[20,8190,8191,8192,8195,8196,8199,8200,8203],{},"To work with LangChain you need to install ",[6150,8193,8194],{},"langchain"," package, as it is not a dependence of Pathway. In the example in this guide you will also use ",[6150,8197,8198],{},"OpenAIEmbeddings"," class for which you need ",[6150,8201,8202],{},"langchain_openai"," package.",[6525,8205,8207],{"className":6527,"code":8206,"language":6529,"meta":201,"style":201},"!pip install langchain\n!pip install langchain_community\n!pip install langchain_openai\n\n",[6150,8208,8209,8214,8219],{"__ignoreMap":201},[392,8210,8211],{"class":6534,"line":6535},[392,8212,8213],{"class":6538},"!pip install langchain\n",[392,8215,8216],{"class":6534,"line":510},[392,8217,8218],{"class":6538},"!pip install langchain_community\n",[392,8220,8221],{"class":6534,"line":897},[392,8222,8223],{"class":6538},"!pip install langchain_openai\n",[15,8225,8227],{"id":8226},"using-langchain-components-in-pathway-vector-store","Using LangChain components in Pathway Vector Store",[20,8229,8230,8231,8237,8238,8244],{},"When using Pathway ",[717,8232,8234],{"href":8233},"/developers/api-docs/pathway-xpacks-llm/vectorstore#pathway.xpacks.llm.vector_store.VectorStoreServer",[6150,8235,8236],{},"VectorStoreServer",", you can use LangChain embedder and splitter for processing documents. To do that, use ",[717,8239,8241],{"href":8240},"/developers/api-docs/pathway-xpacks-llm/vectorstore#pathway.xpacks.llm.vector_store.VectorStoreServer.from_langchain_components",[6150,8242,8243],{},"from_langchain_components"," class method.",[20,8246,8247,8248,3267],{},"To start, you need to create a folder Pathway will listen to. Feel free to skip this if you already have a folder on which you want to build your RAG application. You can also use Google Drive, Sharepoint, or any other source from ",[717,8249,8251],{"href":8250},"/developers/api-docs/pathway-io","pathway-io",[6525,8253,8255],{"className":6527,"code":8254,"language":6529,"meta":201,"style":201},"!mkdir -p 'data/'\n",[6150,8256,8257],{"__ignoreMap":201},[392,8258,8259,8261,8263,8266,8268,8271],{"class":6534,"line":6535},[392,8260,7052],{"class":6538},[392,8262,7055],{"class":6542},[392,8264,8265],{"class":6538},"p ",[392,8267,6543],{"class":6542},[392,8269,8270],{"class":6546},"data/",[392,8272,8273],{"class":6542},"'\n",[20,8275,8276],{},"To run this example you also need to set OpenAI API key, or change the embedder.",[6525,8278,8280],{"className":6527,"code":8279,"language":6529,"meta":201,"style":201},"import os\nimport getpass\n\n# Set OpenAI API Key\nif \"OPENAI_API_KEY\" in os.environ:\n    api_key = os.environ[\"OPENAI_API_KEY\"]\nelse:\n    api_key = getpass.getpass(\"OpenAI API Key:\")\n",[6150,8281,8282,8288,8295,8299,8304,8329,8353,8360],{"__ignoreMap":201},[392,8283,8284,8286],{"class":6534,"line":6535},[392,8285,6583],{"class":6582},[392,8287,6593],{"class":6538},[392,8289,8290,8292],{"class":6534,"line":510},[392,8291,6583],{"class":6582},[392,8293,8294],{"class":6538}," getpass\n",[392,8296,8297],{"class":6534,"line":897},[392,8298,6633],{"emptyLinePlaceholder":931},[392,8300,8301],{"class":6534,"line":6649},[392,8302,8303],{"class":6779},"# Set OpenAI API Key\n",[392,8305,8306,8309,8311,8314,8316,8319,8322,8324,8326],{"class":6534,"line":6654},[392,8307,8308],{"class":6582},"if",[392,8310,6811],{"class":6542},[392,8312,8313],{"class":6546},"OPENAI_API_KEY",[392,8315,6797],{"class":6542},[392,8317,8318],{"class":6542}," in",[392,8320,8321],{"class":6538}," os",[392,8323,3267],{"class":6542},[392,8325,6791],{"class":6564},[392,8327,8328],{"class":6542},":\n",[392,8330,8331,8334,8336,8338,8340,8342,8344,8346,8348,8350],{"class":6534,"line":6679},[392,8332,8333],{"class":6538},"    api_key ",[392,8335,6844],{"class":6542},[392,8337,8321],{"class":6538},[392,8339,3267],{"class":6542},[392,8341,6791],{"class":6564},[392,8343,6794],{"class":6542},[392,8345,6797],{"class":6542},[392,8347,8313],{"class":6546},[392,8349,6797],{"class":6542},[392,8351,8352],{"class":6542},"]\n",[392,8354,8355,8358],{"class":6534,"line":6721},[392,8356,8357],{"class":6582},"else",[392,8359,8328],{"class":6542},[392,8361,8362,8364,8366,8369,8371,8374,8376,8378,8381,8383],{"class":6534,"line":6746},[392,8363,8333],{"class":6538},[392,8365,6844],{"class":6542},[392,8367,8368],{"class":6538}," getpass",[392,8370,3267],{"class":6542},[392,8372,8373],{"class":6918},"getpass",[392,8375,6922],{"class":6542},[392,8377,6797],{"class":6542},[392,8379,8380],{"class":6546},"OpenAI API Key:",[392,8382,6797],{"class":6542},[392,8384,6933],{"class":6542},[20,8386,8387,8388,8390],{},"To run the server use Pathway filesystem connector to read files from the ",[6150,8389,7041],{}," folder.",[6525,8392,8394],{"className":6527,"code":8393,"language":6529,"meta":201,"style":201},"import pathway as pw\n\nfrom pathway.xpacks.llm.vector_store import VectorStoreServer\nfrom langchain_openai import OpenAIEmbeddings\nfrom langchain.text_splitter import CharacterTextSplitter\n\ndata = pw.io.fs.read(\n    \"./data\",\n    format=\"binary\",\n    mode=\"streaming\",\n    with_metadata=True,\n)\n",[6150,8395,8396,8406,8410,8432,8444,8461,8465,8488,8498,8512,8528,8534],{"__ignoreMap":201},[392,8397,8398,8400,8402,8404],{"class":6534,"line":6535},[392,8399,6583],{"class":6582},[392,8401,6640],{"class":6538},[392,8403,6643],{"class":6582},[392,8405,6646],{"class":6538},[392,8407,8408],{"class":6534,"line":510},[392,8409,6633],{"emptyLinePlaceholder":931},[392,8411,8412,8414,8416,8418,8420,8422,8424,8426,8428,8430],{"class":6534,"line":897},[392,8413,6657],{"class":6582},[392,8415,6660],{"class":6538},[392,8417,3267],{"class":6542},[392,8419,6688],{"class":6538},[392,8421,3267],{"class":6542},[392,8423,982],{"class":6538},[392,8425,3267],{"class":6542},[392,8427,6763],{"class":6538},[392,8429,6583],{"class":6582},[392,8431,6768],{"class":6538},[392,8433,8434,8436,8439,8441],{"class":6534,"line":6649},[392,8435,6657],{"class":6582},[392,8437,8438],{"class":6538}," langchain_openai ",[392,8440,6583],{"class":6582},[392,8442,8443],{"class":6538}," OpenAIEmbeddings\n",[392,8445,8446,8448,8451,8453,8456,8458],{"class":6534,"line":6654},[392,8447,6657],{"class":6582},[392,8449,8450],{"class":6538}," langchain",[392,8452,3267],{"class":6542},[392,8454,8455],{"class":6538},"text_splitter ",[392,8457,6583],{"class":6582},[392,8459,8460],{"class":6538}," CharacterTextSplitter\n",[392,8462,8463],{"class":6534,"line":6679},[392,8464,6633],{"emptyLinePlaceholder":931},[392,8466,8467,8470,8472,8474,8476,8478,8480,8482,8484,8486],{"class":6534,"line":6721},[392,8468,8469],{"class":6538},"data ",[392,8471,6844],{"class":6542},[392,8473,7179],{"class":6538},[392,8475,3267],{"class":6542},[392,8477,7184],{"class":6564},[392,8479,3267],{"class":6542},[392,8481,7189],{"class":6564},[392,8483,3267],{"class":6542},[392,8485,7194],{"class":6918},[392,8487,6980],{"class":6542},[392,8489,8490,8492,8494,8496],{"class":6534,"line":6746},[392,8491,7488],{"class":6542},[392,8493,7034],{"class":6546},[392,8495,6797],{"class":6542},[392,8497,7213],{"class":6542},[392,8499,8500,8502,8504,8506,8508,8510],{"class":6534,"line":6771},[392,8501,7218],{"class":6925},[392,8503,6844],{"class":6542},[392,8505,6797],{"class":6542},[392,8507,7225],{"class":6546},[392,8509,6797],{"class":6542},[392,8511,7213],{"class":6542},[392,8513,8514,8517,8519,8521,8524,8526],{"class":6534,"line":6776},[392,8515,8516],{"class":6925},"    mode",[392,8518,6844],{"class":6542},[392,8520,6797],{"class":6542},[392,8522,8523],{"class":6546},"streaming",[392,8525,6797],{"class":6542},[392,8527,7213],{"class":6542},[392,8529,8530,8532],{"class":6534,"line":6783},[392,8531,7234],{"class":6925},[392,8533,7237],{"class":6542},[392,8535,8536],{"class":6534,"line":6983},[392,8537,6933],{"class":6542},[20,8539,8540,8541,8544,8545,8547],{},"And then pass them to the server, which will split them using ",[6150,8542,8543],{},"CharacterTextSplitter"," and embed them using ",[6150,8546,8198],{},", both from LangChain.",[6525,8549,8551],{"className":6527,"code":8550,"language":6529,"meta":201,"style":201},"embeddings = OpenAIEmbeddings(api_key=api_key)\nsplitter = CharacterTextSplitter()\n\nhost = \"127.0.0.1\"\nport = 8666\n\nserver = VectorStoreServer.from_langchain_components(\n    data, embedder=embeddings, splitter=splitter\n)\nserver.run_server(host, port=port, with_cache=True, cache_backend=pw.persistence.Backend.filesystem(\"./Cache\"), threaded=True)\n",[6150,8552,8553,8573,8585,8589,8603,8613,8617,8632,8657,8661],{"__ignoreMap":201},[392,8554,8555,8558,8560,8563,8565,8567,8569,8571],{"class":6534,"line":6535},[392,8556,8557],{"class":6538},"embeddings ",[392,8559,6844],{"class":6542},[392,8561,8562],{"class":6918}," OpenAIEmbeddings",[392,8564,6922],{"class":6542},[392,8566,6926],{"class":6925},[392,8568,6844],{"class":6542},[392,8570,6926],{"class":6918},[392,8572,6933],{"class":6542},[392,8574,8575,8578,8580,8583],{"class":6534,"line":510},[392,8576,8577],{"class":6538},"splitter ",[392,8579,6844],{"class":6542},[392,8581,8582],{"class":6918}," CharacterTextSplitter",[392,8584,7976],{"class":6542},[392,8586,8587],{"class":6534,"line":897},[392,8588,6633],{"emptyLinePlaceholder":931},[392,8590,8591,8594,8596,8598,8601],{"class":6534,"line":6649},[392,8592,8593],{"class":6538},"host ",[392,8595,6844],{"class":6542},[392,8597,6811],{"class":6542},[392,8599,8600],{"class":6546},"127.0.0.1",[392,8602,6852],{"class":6542},[392,8604,8605,8608,8610],{"class":6534,"line":6654},[392,8606,8607],{"class":6538},"port ",[392,8609,6844],{"class":6542},[392,8611,8612],{"class":6558}," 8666\n",[392,8614,8615],{"class":6534,"line":6679},[392,8616,6633],{"emptyLinePlaceholder":931},[392,8618,8619,8622,8624,8626,8628,8630],{"class":6534,"line":6721},[392,8620,8621],{"class":6538},"server ",[392,8623,6844],{"class":6542},[392,8625,7681],{"class":6538},[392,8627,3267],{"class":6542},[392,8629,8243],{"class":6918},[392,8631,6980],{"class":6542},[392,8633,8634,8637,8639,8642,8644,8647,8649,8652,8654],{"class":6534,"line":6746},[392,8635,8636],{"class":6918},"    data",[392,8638,6673],{"class":6542},[392,8640,8641],{"class":6925}," embedder",[392,8643,6844],{"class":6542},[392,8645,8646],{"class":6918},"embeddings",[392,8648,6673],{"class":6542},[392,8650,8651],{"class":6925}," splitter",[392,8653,6844],{"class":6542},[392,8655,8656],{"class":6918},"splitter\n",[392,8658,8659],{"class":6534,"line":6771},[392,8660,6933],{"class":6542},[392,8662,8663,8666,8668,8670,8672,8674,8676,8678,8680,8683,8685,8688,8691,8694,8696,8698,8700,8703,8705,8708,8710,8713,8715,8717,8720,8722,8725,8728],{"class":6534,"line":6776},[392,8664,8665],{"class":6538},"server",[392,8667,3267],{"class":6542},[392,8669,7927],{"class":6918},[392,8671,6922],{"class":6542},[392,8673,7869],{"class":6918},[392,8675,6673],{"class":6542},[392,8677,7879],{"class":6925},[392,8679,6844],{"class":6542},[392,8681,8682],{"class":6918},"port",[392,8684,6673],{"class":6542},[392,8686,8687],{"class":6925}," with_cache",[392,8689,8690],{"class":6542},"=True,",[392,8692,8693],{"class":6925}," cache_backend",[392,8695,6844],{"class":6542},[392,8697,6947],{"class":6918},[392,8699,3267],{"class":6542},[392,8701,8702],{"class":6564},"persistence",[392,8704,3267],{"class":6542},[392,8706,8707],{"class":6564},"Backend",[392,8709,3267],{"class":6542},[392,8711,8712],{"class":6918},"filesystem",[392,8714,6922],{"class":6542},[392,8716,6797],{"class":6542},[392,8718,8719],{"class":6546},"./Cache",[392,8721,6797],{"class":6542},[392,8723,8724],{"class":6542},"),",[392,8726,8727],{"class":6925}," threaded",[392,8729,8730],{"class":6542},"=True)\n",[20,8732,8733,8734,8739,8740,8742,8743,8746],{},"The server is now running and ready for querying with a ",[717,8735,8737],{"href":8736},"/developers/api-docs/pathway-xpacks-llm/vectorstore#pathway.xpacks.llm.vector_store.VectorStoreClient",[6150,8738,8236],{}," or with a ",[6150,8741,8181],{}," from ",[6150,8744,8745],{},"langchain-community"," described in the next Section.",[15,8748,8750],{"id":8749},"using-pathway-as-a-vector-store-in-langchain-pipelines","Using Pathway as a Vector Store in LangChain pipelines",[20,8752,8753,8754,8756,8757,3267],{},"Once you have a ",[6150,8755,8236],{}," running you can access it from LangChain pipeline by using ",[717,8758,8181],{"href":8179,"rel":8759},[733],[20,8761,8762,8763,4426,8766,4130,8768,8770,8771,8773,8774,8776,8777,8782,8783,8786,8787,4130,8792,3267],{},"To do that you need to provide either the ",[6150,8764,8765],{},"url",[6150,8767,7869],{},[6150,8769,8682],{}," of the running ",[6150,8772,8236],{},". In the code example below, you will connect to the ",[6150,8775,8236],{}," defined in the previous Section, so make sure it's running before making queries. Alternatively, you can also use a publicly available ",[717,8778,8781],{"href":8779,"rel":8780},"https://pathway.com/solutions",[733],"demo pipeline"," to test your client. Its REST API you can access at ",[6150,8784,8785],{},"https://demo-document-indexing.pathway.stream",". This demo ingests documents from ",[717,8788,8791],{"href":8789,"rel":8790},"https://drive.google.com/drive/u/0/folders/1cULDv2OaViJBmOfG5WB0oWcgayNrGtVs",[733],"Google Drive",[717,8793,8796],{"href":8794,"rel":8795},"https://navalgo.sharepoint.com/sites/ConnectorSandbox/Shared%20Documents/Forms/AllItems.aspx?id=%2Fsites%2FConnectorSandbox%2FShared%20Documents%2FIndexerSandbox&p=true&ga=1",[733],"Sharepoint",[6525,8798,8800],{"className":6527,"code":8799,"language":6529,"meta":201,"style":201},"from langchain_community.vectorstores import PathwayVectorClient\n\nclient = PathwayVectorClient(host=host, port=port)\n",[6150,8801,8802,8819,8823],{"__ignoreMap":201},[392,8803,8804,8806,8809,8811,8814,8816],{"class":6534,"line":6535},[392,8805,6657],{"class":6582},[392,8807,8808],{"class":6538}," langchain_community",[392,8810,3267],{"class":6542},[392,8812,8813],{"class":6538},"vectorstores ",[392,8815,6583],{"class":6582},[392,8817,8818],{"class":6538}," PathwayVectorClient\n",[392,8820,8821],{"class":6534,"line":510},[392,8822,6633],{"emptyLinePlaceholder":931},[392,8824,8825,8827,8829,8832,8834,8836,8838,8840,8842,8844,8846,8848],{"class":6534,"line":897},[392,8826,8018],{"class":6538},[392,8828,6844],{"class":6542},[392,8830,8831],{"class":6918}," PathwayVectorClient",[392,8833,6922],{"class":6542},[392,8835,7869],{"class":6925},[392,8837,6844],{"class":6542},[392,8839,7869],{"class":6918},[392,8841,6673],{"class":6542},[392,8843,7879],{"class":6925},[392,8845,6844],{"class":6542},[392,8847,8682],{"class":6918},[392,8849,6933],{"class":6542},[6525,8851,8853],{"className":6527,"code":8852,"language":6529,"meta":201,"style":201},"query = \"What is Pathway?\"\ndocs = client.similarity_search(query)\nprint(docs)\n",[6150,8854,8855,8869,8890],{"__ignoreMap":201},[392,8856,8857,8860,8862,8864,8867],{"class":6534,"line":6535},[392,8858,8859],{"class":6538},"query ",[392,8861,6844],{"class":6542},[392,8863,6811],{"class":6542},[392,8865,8866],{"class":6546},"What is Pathway?",[392,8868,6852],{"class":6542},[392,8870,8871,8874,8876,8878,8880,8883,8885,8888],{"class":6534,"line":510},[392,8872,8873],{"class":6538},"docs ",[392,8875,6844],{"class":6542},[392,8877,8070],{"class":6538},[392,8879,3267],{"class":6542},[392,8881,8882],{"class":6918},"similarity_search",[392,8884,6922],{"class":6542},[392,8886,8887],{"class":6918},"query",[392,8889,6933],{"class":6542},[392,8891,8892,8894,8896,8899],{"class":6534,"line":897},[392,8893,8091],{"class":6918},[392,8895,6922],{"class":6542},[392,8897,8898],{"class":6918},"docs",[392,8900,6933],{"class":6542},[20,8902,8903],{},"As you can see, the LLM cannot respond clearly as it lacks current knowledge, but this is where Pathway shines. Add new data to the folder Pathway is listening to, then ask our agent again to see how it responds.\nTo do that, you can download the repo readme of Pathway into our data folder:",[6525,8905,8907],{"className":6527,"code":8906,"language":6529,"meta":201,"style":201},"!wget 'https://raw.githubusercontent.com/pathwaycom/pathway/main/README.md' -O 'data/pathway_readme.md' -q -nc\n",[6150,8908,8909],{"__ignoreMap":201},[392,8910,8911,8913,8915,8918,8920,8923,8926,8928,8931,8933,8935,8937,8939],{"class":6534,"line":6535},[392,8912,7073],{"class":6538},[392,8914,6543],{"class":6542},[392,8916,8917],{"class":6546},"https://raw.githubusercontent.com/pathwaycom/pathway/main/README.md",[392,8919,6543],{"class":6542},[392,8921,8922],{"class":6542}," -",[392,8924,8925],{"class":6538},"O ",[392,8927,6543],{"class":6542},[392,8929,8930],{"class":6546},"data/pathway_readme.md",[392,8932,6543],{"class":6542},[392,8934,8922],{"class":6542},[392,8936,7078],{"class":6538},[392,8938,7055],{"class":6542},[392,8940,8941],{"class":6538},"nc\n",[20,8943,8944],{},"Try again to query with the new data:",[6525,8946,8948],{"className":6527,"code":8947,"language":6529,"meta":201,"style":201},"docs = client.similarity_search(query)\nprint(docs)\n",[6150,8949,8950,8968],{"__ignoreMap":201},[392,8951,8952,8954,8956,8958,8960,8962,8964,8966],{"class":6534,"line":6535},[392,8953,8873],{"class":6538},[392,8955,6844],{"class":6542},[392,8957,8070],{"class":6538},[392,8959,3267],{"class":6542},[392,8961,8882],{"class":6918},[392,8963,6922],{"class":6542},[392,8965,8887],{"class":6918},[392,8967,6933],{"class":6542},[392,8969,8970,8972,8974,8976],{"class":6534,"line":510},[392,8971,8091],{"class":6918},[392,8973,6922],{"class":6542},[392,8975,8898],{"class":6918},[392,8977,6933],{"class":6542},[64,8979,8981],{"id":8980},"rag-pipeline-in-langchain","RAG pipeline in LangChain",[20,8983,8984],{},"The next step is to write a chain in LangChain. The next example implements a simple RAG, that given a question, retrieves documents from Pathway Vector Store. These are then used as a context for the given question in a prompt sent to the OpenAI chat.",[6525,8986,8988],{"className":6527,"code":8987,"language":6529,"meta":201,"style":201},"from langchain_core.output_parsers import StrOutputParser\nfrom langchain_core.prompts import ChatPromptTemplate\nfrom langchain_core.runnables import RunnablePassthrough\nfrom langchain_openai import ChatOpenAI\n\nretriever = client.as_retriever()\n\ntemplate = \"\"\"\nYou are smart assistant that helps users with their documents on Google Drive and Sharepoint.\nGiven a context, respond to the user question.\nCONTEXT:\n{context}\nQUESTION: {question}\nYOUR ANSWER:\"\"\"\n\nprompt = ChatPromptTemplate.from_template(template)\nllm = ChatOpenAI()\nchain = (\n    {\"context\": retriever, \"question\": RunnablePassthrough()}\n    | prompt\n    | llm\n    | StrOutputParser()\n)\n",[6150,8989,8990,9007,9023,9039,9050,9054,9070,9074,9084,9089,9094,9099,9104,9112,9120,9125,9147,9159,9170,9205,9214,9222,9232],{"__ignoreMap":201},[392,8991,8992,8994,8997,8999,9002,9004],{"class":6534,"line":6535},[392,8993,6657],{"class":6582},[392,8995,8996],{"class":6538}," langchain_core",[392,8998,3267],{"class":6542},[392,9000,9001],{"class":6538},"output_parsers ",[392,9003,6583],{"class":6582},[392,9005,9006],{"class":6538}," StrOutputParser\n",[392,9008,9009,9011,9013,9015,9018,9020],{"class":6534,"line":510},[392,9010,6657],{"class":6582},[392,9012,8996],{"class":6538},[392,9014,3267],{"class":6542},[392,9016,9017],{"class":6538},"prompts ",[392,9019,6583],{"class":6582},[392,9021,9022],{"class":6538}," ChatPromptTemplate\n",[392,9024,9025,9027,9029,9031,9034,9036],{"class":6534,"line":897},[392,9026,6657],{"class":6582},[392,9028,8996],{"class":6538},[392,9030,3267],{"class":6542},[392,9032,9033],{"class":6538},"runnables ",[392,9035,6583],{"class":6582},[392,9037,9038],{"class":6538}," RunnablePassthrough\n",[392,9040,9041,9043,9045,9047],{"class":6534,"line":6649},[392,9042,6657],{"class":6582},[392,9044,8438],{"class":6538},[392,9046,6583],{"class":6582},[392,9048,9049],{"class":6538}," ChatOpenAI\n",[392,9051,9052],{"class":6534,"line":6654},[392,9053,6633],{"emptyLinePlaceholder":931},[392,9055,9056,9059,9061,9063,9065,9068],{"class":6534,"line":6679},[392,9057,9058],{"class":6538},"retriever ",[392,9060,6844],{"class":6542},[392,9062,8070],{"class":6538},[392,9064,3267],{"class":6542},[392,9066,9067],{"class":6918},"as_retriever",[392,9069,7976],{"class":6542},[392,9071,9072],{"class":6534,"line":6721},[392,9073,6633],{"emptyLinePlaceholder":931},[392,9075,9076,9079,9081],{"class":6534,"line":6746},[392,9077,9078],{"class":6538},"template ",[392,9080,6844],{"class":6542},[392,9082,9083],{"class":6542}," \"\"\"\n",[392,9085,9086],{"class":6534,"line":6771},[392,9087,9088],{"class":6546},"You are smart assistant that helps users with their documents on Google Drive and Sharepoint.\n",[392,9090,9091],{"class":6534,"line":6776},[392,9092,9093],{"class":6546},"Given a context, respond to the user question.\n",[392,9095,9096],{"class":6534,"line":6783},[392,9097,9098],{"class":6546},"CONTEXT:\n",[392,9100,9101],{"class":6534,"line":6983},[392,9102,9103],{"class":6558},"{context}\n",[392,9105,9106,9109],{"class":6534,"line":7022},[392,9107,9108],{"class":6546},"QUESTION: ",[392,9110,9111],{"class":6558},"{question}\n",[392,9113,9114,9117],{"class":6534,"line":7625},[392,9115,9116],{"class":6546},"YOUR ANSWER:",[392,9118,9119],{"class":6542},"\"\"\"\n",[392,9121,9123],{"class":6534,"line":9122},15,[392,9124,6633],{"emptyLinePlaceholder":931},[392,9126,9128,9131,9133,9136,9138,9141,9143,9145],{"class":6534,"line":9127},16,[392,9129,9130],{"class":6538},"prompt ",[392,9132,6844],{"class":6542},[392,9134,9135],{"class":6538}," ChatPromptTemplate",[392,9137,3267],{"class":6542},[392,9139,9140],{"class":6918},"from_template",[392,9142,6922],{"class":6542},[392,9144,751],{"class":6918},[392,9146,6933],{"class":6542},[392,9148,9150,9152,9154,9157],{"class":6534,"line":9149},17,[392,9151,6693],{"class":6538},[392,9153,6844],{"class":6542},[392,9155,9156],{"class":6918}," ChatOpenAI",[392,9158,7976],{"class":6542},[392,9160,9162,9165,9167],{"class":6534,"line":9161},18,[392,9163,9164],{"class":6538},"chain ",[392,9166,6844],{"class":6542},[392,9168,9169],{"class":6542}," (\n",[392,9171,9173,9176,9178,9181,9183,9185,9188,9190,9192,9195,9197,9199,9202],{"class":6534,"line":9172},19,[392,9174,9175],{"class":6542},"    {",[392,9177,6797],{"class":6542},[392,9179,9180],{"class":6546},"context",[392,9182,6797],{"class":6542},[392,9184,7496],{"class":6542},[392,9186,9187],{"class":6538}," retriever",[392,9189,6673],{"class":6542},[392,9191,6811],{"class":6542},[392,9193,9194],{"class":6546},"question",[392,9196,6797],{"class":6542},[392,9198,7496],{"class":6542},[392,9200,9201],{"class":6918}," RunnablePassthrough",[392,9203,9204],{"class":6542},"()}\n",[392,9206,9208,9211],{"class":6534,"line":9207},20,[392,9209,9210],{"class":6542},"    |",[392,9212,9213],{"class":6538}," prompt\n",[392,9215,9217,9219],{"class":6534,"line":9216},21,[392,9218,9210],{"class":6542},[392,9220,9221],{"class":6538}," llm\n",[392,9223,9225,9227,9230],{"class":6534,"line":9224},22,[392,9226,9210],{"class":6542},[392,9228,9229],{"class":6918}," StrOutputParser",[392,9231,7976],{"class":6542},[392,9233,9235],{"class":6534,"line":9234},23,[392,9236,6933],{"class":6542},[20,9238,9239],{},"Now you have a RAG chain written in LangChain that uses Pathway as its Vector Store. Test it by asking some question.",[6525,9241,9243],{"className":6527,"code":9242,"language":6529,"meta":201,"style":201},"chain.invoke(\"What is Pathway?\")\n",[6150,9244,9245],{"__ignoreMap":201},[392,9246,9247,9250,9252,9255,9257,9259,9261,9263],{"class":6534,"line":6535},[392,9248,9249],{"class":6538},"chain",[392,9251,3267],{"class":6542},[392,9253,9254],{"class":6918},"invoke",[392,9256,6922],{"class":6542},[392,9258,6797],{"class":6542},[392,9260,8866],{"class":6546},[392,9262,6797],{"class":6542},[392,9264,6933],{"class":6542},[64,9266,9268],{"id":9267},"vector-store-statistics","Vector Store statistics",[20,9270,9271,9272,9277,9278,9280],{},"Just like ",[717,9273,9274],{"href":8736},[6150,9275,9276],{},"VectorStoreClient"," from the Pathway LLM xpack, ",[6150,9279,8181],{}," gives you two methods for getting information about indexed documents.",[20,9282,9283,9284,9291,9292,9299],{},"The first one is ",[717,9285,9288],{"href":9286,"rel":9287},"https://api.python.langchain.com/en/latest/vectorstores/langchain_community.vectorstores.pathway.PathwayVectorClient.html#langchain_community.vectorstores.pathway.PathwayVectorClient.get_vectorstore_statistics",[733],[6150,9289,9290],{},"get_vectorstore_statistics"," and gives essential statistics on the state of the vector store, like the number of indexed files and the timestamp of the last updated one. The second one is ",[717,9293,9296],{"href":9294,"rel":9295},"https://api.python.langchain.com/en/latest/vectorstores/langchain_community.vectorstores.pathway.PathwayVectorClient.html#langchain_community.vectorstores.pathway.PathwayVectorClient.get_input_files",[733],[6150,9297,9298],{},"get_input_files",", which gets the list of indexed files along with the associated metadata.",[6525,9301,9303],{"className":6527,"code":9302,"language":6529,"meta":201,"style":201},"print(client.get_vectorstore_statistics())\nprint(client.get_input_files())\n",[6150,9304,9305,9321],{"__ignoreMap":201},[392,9306,9307,9309,9311,9314,9316,9318],{"class":6534,"line":6535},[392,9308,8091],{"class":6918},[392,9310,6922],{"class":6542},[392,9312,9313],{"class":6918},"client",[392,9315,3267],{"class":6542},[392,9317,9290],{"class":6918},[392,9319,9320],{"class":6542},"())\n",[392,9322,9323,9325,9327,9329,9331,9333],{"class":6534,"line":510},[392,9324,8091],{"class":6918},[392,9326,6922],{"class":6542},[392,9328,9313],{"class":6918},[392,9330,3267],{"class":6542},[392,9332,9298],{"class":6918},[392,9334,9320],{"class":6542},[8122,9336,9337],{},"html pre.shiki code .s0W1g, html code.shiki .s0W1g{--shiki-default:#BABED8}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html pre.shiki code .sAklC, html code.shiki .sAklC{--shiki-default:#89DDFF}html pre.shiki code .sfyAc, html code.shiki .sfyAc{--shiki-default:#C3E88D}html pre.shiki code .s6cf3, html code.shiki .s6cf3{--shiki-default:#89DDFF;--shiki-default-font-style:italic}html pre.shiki code .saEQR, html code.shiki .saEQR{--shiki-default:#676E95;--shiki-default-font-style:italic}html pre.shiki code .s-wAU, html code.shiki .s-wAU{--shiki-default:#F07178}html pre.shiki code .sdLwU, html code.shiki .sdLwU{--shiki-default:#82AAFF}html pre.shiki code .s7ZW3, html code.shiki .s7ZW3{--shiki-default:#BABED8;--shiki-default-font-style:italic}html pre.shiki code .sx098, html code.shiki .sx098{--shiki-default:#F78C6C}",{"title":201,"searchDepth":510,"depth":510,"links":9339},[9340,9341,9342],{"id":8187,"depth":510,"text":8188},{"id":8226,"depth":510,"text":8227},{"id":8749,"depth":510,"text":8750,"children":9343},[9344,9345],{"id":8980,"depth":897,"text":8981},{"id":9267,"depth":897,"text":9268},{"layout":925,"date":1002,"hidden":931,"thumbnail":9347,"tags":9348,"notebook_export_path":1007,"keywords":9349},{"src":1004},[981,982,1006],[984,985,986,987,1009,992],{"title":998,"description":201},{"loc":999},"bg_TqVKqcxb1uD2Bj0KQcx_-HRvp_C3sMuvhQGF-dGA",[9354,9355],{"title":1782,"path":1783,"stem":1784,"children":-1},{"title":1795,"path":1796,"stem":1797,"children":-1},1775364309663]