--- title: Open Source Extraction Service | daily.dev description: LangChain has launched an open-source extraction service that uses LLMs to pull structured data from unstructured sources like PDFs, HTML, and text. The hosted... canonical: https://daily.dev/posts/open-source-extraction-service-v9yijohcv twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Open Source Extraction Service | daily.dev og:description: LangChain has launched an open-source extraction service that uses LLMs to pull structured data from unstructured sources like PDFs, HTML, and text. The hosted... og:url: https://daily.dev/posts/open-source-extraction-service-v9yijohcv og:image: https://api.daily.dev/og/posts/V9yiJohcV.png og:image:alt: Open Source Extraction Service og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Open Source Extraction Service **[LangChain](https://daily.dev/sources/langchain)** · 8 min read · 0 upvotes · 0 comments ## Summary LangChain has launched an open-source extraction service that uses LLMs to pull structured data from unstructured sources like PDFs, HTML, and text. The hosted demo is free for experimentation. Key features include custom schema definition, few-shot examples for in-context learning, support for multiple LLM models, extractor sharing, and a LangServe endpoint for integration into larger LangChain workflows. A detailed walkthrough demonstrates extracting financial figures from Uber's Q4 2023 earnings call transcript using Pydantic schemas and the Python API, including how few-shot examples improve output formatting consistency. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How do I extract structured data like financial figures from a PDF using LangChain? Define a Pydantic schema describing the fields to extract (e.g., name, value, scale, period), post it to LangChain's open-source extraction service to create an extractor, then send the PDF bytes to the /extract endpoint along with a model name like gpt-3.5-turbo. The service returns JSON records with an evidence field showing the source sentence for each extracted value. _daily.dev surfaces practical LLM extraction workflows for developers building similar data pipelines._ ### How can few-shot examples improve LLM data extraction accuracy with LangChain's extraction service? Adding a single few-shot example with the desired output formatting (e.g., using 'MM' instead of 'million' for scale, or lowercase field names) corrected inconsistent LLM outputs when extracting financial data from Uber's Q4 2023 earnings transcript using gpt-3.5-turbo. This targeted fix avoided needing to switch to a larger, more expensive model to achieve the intended formatting. _developers tuning LLM prompts for structured output track techniques like this via daily.dev._ ### How do I integrate a LangServe extraction endpoint into a retrieval-augmented workflow? Use the RemoteRunnable client from the langserve package to invoke a hosted extraction endpoint, then chain it with a retriever (such as FAISS) so that a search query first retrieves the relevant document content before passing it to the extraction schema. This lets a query like 'rev' fetch a matching document and restrict extraction to just that content. _engineers building agent and RAG pipelines follow integration patterns like this on daily.dev._ ## Similar posts on daily.dev - [Automating Web Research](https://daily.dev/posts/automating-web-research-pwuhcbqf7) · LangChain · 0 upvotes · 0 comments - [Untitled](https://daily.dev/posts/untitled-la53kergy) · SitePoint · 0 upvotes · 0 comments - [Automating Knowledge Graph Population: Extracting Entities and Triples from Unstructured Text with an LLM](https://daily.dev/posts/automating-knowledge-graph-population-extracting-entities-and-triples-from-unstructured-text-with-a-occ2gvqzy) · Machine Learning Mastery · 1 upvotes · 0 comments - [Medium](https://daily.dev/posts/medium-wrdoj9xmv) · Medium · 2 upvotes · 0 comments - [Cube x LangChain: Building AI experiences with LLMs and the semantic layer](https://daily.dev/posts/cube-x-langchain-building-ai-experiences-with-llms-and-the-semantic-layer-bqmdxwkbo) · LangChain · 0 upvotes · 0 comments --- Tags: [#python](https://daily.dev/tags/python), [#llm](https://daily.dev/tags/llm), [#langchain](https://daily.dev/tags/langchain), [#pydantic](https://daily.dev/tags/pydantic) [View this post on daily.dev](https://daily.dev/posts/open-source-extraction-service-v9yijohcv) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Open Source Extraction Service","url":"https://daily.dev/posts/open-source-extraction-service-v9yijohcv","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/open-source-extraction-service-v9yijohcv"},"datePublished":"2026-07-18T01:08:29.169Z","dateModified":"2026-09-13T20:29:32.022Z","description":"LangChain has launched an open-source extraction service that uses LLMs to pull structured data from unstructured sources like PDFs, HTML, and text. The hosted...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6ff482b9f30f982d7a1a36018b85b33b?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/6ff482b9f30f982d7a1a36018b85b33b?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"LangChain","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"LangChain","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/0f4f2629e0724a849823f8cd0d913e13","url":"https://daily.dev/sources/langchain"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/open-source-extraction-service-v9yijohcv","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"python,llm,langchain,pydantic","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"LangChain","item":"https://daily.dev/sources/langchain"},{"@type":"ListItem","position":3,"name":"Open Source Extraction Service"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/open-source-extraction-service-v9yijohcv#faq","mainEntity":[{"@type":"Question","name":"How do I extract structured data like financial figures from a PDF using LangChain?","acceptedAnswer":{"@type":"Answer","text":"Define a Pydantic schema describing the fields to extract (e.g., name, value, scale, period), post it to LangChain's open-source extraction service to create an extractor, then send the PDF bytes to the /extract endpoint along with a model name like gpt-3.5-turbo. The service returns JSON records with an evidence field showing the source sentence for each extracted value. daily.dev surfaces practical LLM extraction workflows for developers building similar data pipelines."}},{"@type":"Question","name":"How can few-shot examples improve LLM data extraction accuracy with LangChain's extraction service?","acceptedAnswer":{"@type":"Answer","text":"Adding a single few-shot example with the desired output formatting (e.g., using 'MM' instead of 'million' for scale, or lowercase field names) corrected inconsistent LLM outputs when extracting financial data from Uber's Q4 2023 earnings transcript using gpt-3.5-turbo. This targeted fix avoided needing to switch to a larger, more expensive model to achieve the intended formatting. developers tuning LLM prompts for structured output track techniques like this via daily.dev."}},{"@type":"Question","name":"How do I integrate a LangServe extraction endpoint into a retrieval-augmented workflow?","acceptedAnswer":{"@type":"Answer","text":"Use the RemoteRunnable client from the langserve package to invoke a hosted extraction endpoint, then chain it with a retriever (such as FAISS) so that a search query first retrieves the relevant document content before passing it to the extraction schema. This lets a query like 'rev' fetch a matching document and restrict extraction to just that content. engineers building agent and RAG pipelines follow integration patterns like this on daily.dev."}}]} ```