Building an Auto-Insurance Agentic Workflow from Scratch 🚗 This holiday weekend, learn how to build an agentic workflow that can parse auto insurance claims, retrieve and apply relevant policy guidelines and the specific declarations document, and produce a structured, transparent recommendation to both parties on how much the claim is covered. This guide interleaves structured extraction with various types of retrieval (chunks for policy guidelines, entire files for declarations), to produce the final output. Notebook: https://lnkd.in/g8EpAKRR Signup for LlamaCloud: https://lnkd.in/gi8dxGnt
LlamaIndex
Technology, Information and Internet
San Francisco, California 226,872 followers
The fastest way to build production-quality LLM agents over your data
About us
The data framework for LLMs Python: Github: https://github.com/jerryjliu/llama_index Docs: https://docs.llamaindex.ai/ Typescript/Javascript: Github: https://github.com/run-llama/LlamaIndexTS Docs: https://ts.llamaindex.ai/ Other: Discord: discord.gg/dGcwcsnxhU LlamaHub: llamahub.ai Twitter: https://twitter.com/llama_index Blog: blog.llamaindex.ai #ai #llms #rag
- Website
-
https://www.llamaindex.ai/
External link for LlamaIndex
- Industry
- Technology, Information and Internet
- Company size
- 2-10 employees
- Headquarters
- San Francisco, California
- Type
- Public Company
Locations
-
Primary
Get directions
San Francisco, California, US
Employees at LlamaIndex
Updates
-
See how to use LlamaIndex to automate document processing workflows! In this notebook from Ravi Theja Desetty we show you how to use LlamaIndex to handle a real-world use-case: standardizing units and measurements across multiple vendors and invoices. It's a great example of Workflows, Pydantic structured data extraction, and agentic processing in action: https://lnkd.in/gSmWXN9A
-
-
An early holiday gift for you -- LlamaParse can now parse audio files! LlamaParse is already the world's best parser of complex document formats like PDFs, Word, PowerPoint, spreadsheets and many more, but now it can turn speech into text! Just upload an audio file like you would any other file and we take care of the rest. Check out this intro video from Pierre-Loic Doulcet, or head straight to cloud.llamaindex.ai to try it out yourself!
-
It's been a huge year for LlamaIndex! Tens of millions of pages parsed, millions of framework downloads, and tons of happy users! Check out our year-end review in the numbers, including a month-by-month breakdown of our biggest feature releases: https://lnkd.in/gD83ghUM
-
-
Hanane continues to impress with her in-depth posts, this time on creating a stock analysis bot that works with one click! ➡️ Learn how to build an automated stock analysis agent using LlamaIndex's FunctionCallingAgent combined with Claude 3.5 Sonnet, requiring just one prompt for comprehensive analysis ➡️ See how to integrate a Code Interpreter tool with LlamaIndex to enable dynamic Python execution for tasks like date detection and price fetching, avoiding reliance on the LLM's knowledge cutoff ➡️ Understand the practical implementation pattern of first generating date detection code before fetching historical prices, ensuring analyses stay current and accurate Check out the post below, or go straight to the notebook here: https://lnkd.in/g86Bjea3
🤖 One-Click Stock Analysis: Building an AI Agent with LlamaIndex and Anthropic Claude 3.5 Sonnet Using Code Interpreter I've developed an automated financial analysis agent that leverages LlamaIndex's FunctionCallingAgent architecture integrated with Anthropic's Claude 3.5 Sonnet LLM. This implementation enables programmatic generation of comprehensive stock analysis through a streamlined single-prompt interface. 📈 🎯 What it does: • Automatically detects the current date • Fetches historical stock prices for the current month • Provides detailed statistical analysis and performance metrics • All with a single prompt! 💡 Key Implementation Details: • Built using LlamaIndex's FunctionCallingAgent abstraction • Leverages Code Interpreter tool for seamless Python execution • Powered by Anthropic's Claude 3.5 Sonnet for reliable code generation • Implements smart date detection for accurate historical data fetching 🔍 Sample Results for Tesla (TSLA): • Period: Nov 1-17, 2024 • Total Return: +28.81% • Average Daily Return: 2.76% • Daily Volatility: 6.64% • Price Range: $238.88 - $358.64 🛠️ Tech Stack: ```python from llama_index.tools.code_interpreter.base import CodeInterpreterToolSpec from llama_index.llms.anthropic import Anthropic from llama_index.core.agent import FunctionCallingAgent ``` Tip: First, ask to generate Python code to detect today's date, then request it to fetch historical prices. This ensures your analysis aligns with current dates rather than relying on the LLM's cutoff knowledge, providing consistent and meaningful insights. — Enjoy this? ♻️ Repost it and share it with your network. Want to learn more about building AI-powered tools? Let's connect! Drop your thoughts in the comments below 👇 --- P.S. Check out my GitHub for the complete implementation - link in the comment👇
-
-
Help your community by sharing where you are in your AI journey in Vercel's State of AI Survey: https://lnkd.in/gMKQif2B
-
Learn how to use Vectara's powerful RAG capabilities! 🔍 Discover how to: ➡️ Load data into Vectara ➡️ Query with streaming and reranking options ➡️ Implement chat functionality ➡️ Build agentic RAG applications using vectara-agentic Vectara's end-to-end managed service for RAG includes state-of-the-art embedding, hybrid search, and various reranking strategies. Check out the detailed Jupyter notebook to get started: https://lnkd.in/g3TehD7U
-
-
We believe multi-agent systems are the future of knowledge assistants, and this post goes into detail about how to build one. ➡️ How to evolve from a single agent to a coordinated multi-agent system using LlamaIndex ➡️ Practical code examples for implementing agent factories, system prompts and state management ➡️ Techniques for handling errors and validating data in multi-agent systems ➡️ How to coordinate agent communication and workflow using an Orchestrator agent ➡️ Real-world application through building an Anki flashcard generator with specialized agents We're seeing more an more of this Orchestrator pattern in the wild. Check out the full post: https://lnkd.in/gg-DdnZV
-
-
Learn how to build agents from scratch in this crash course from Tarun R Jain! ➡️ Learn function calling in LlamaIndex to handle real-time data queries using 3 different approaches ➡️ Build agentic RAG that intelligently routes between vector and summary tools ➡️ Create ReAct agents using thought-action-observation loops ➡️ Construct a real-time search assistant as a practical example Check the video here: https://lnkd.in/gyEwyKdB
-
-
Check out this AI-powered SDR that automatically generates leads for your business, built with LlamaIndex! Original tweet here: https://lnkd.in/g4SVFfiq Full code here: https://lnkd.in/ga84_9tY