🤖 Agentic RAG with LlamaIndex

A hands-on tutorial series on building advanced Retrieval-Augmented Generation (RAG) systems using LlamaIndex, with a focus on agentic capabilities and multi-document reasoning.

📚 Description

This repository contains a progressive series of Jupyter notebooks that demonstrate how to build increasingly sophisticated RAG systems using LlamaIndex. The course starts with basic router query engines and progresses to advanced multi-document agents with dynamic tool selection.

Each lesson builds upon the previous one, introducing new concepts and techniques for working with document-based knowledge and LLM agents.

🔍 Features

• Router Query Engines: Learn how to route queries to the most appropriate query engine based on the query content
• Tool Calling: Implement function-calling capabilities to enable LLMs to use tools for retrieving information
• Agent Reasoning Loop: Build an agent that can reason over documents and maintain a conversation
• Multi-Document Agent: Create a sophisticated agent that can work with multiple documents and dynamically select the most relevant tools

🛠️ Prerequisites

• Python 3.9+
• OpenAI API key (stored in a .env file)
• Required packages (see requirements.txt)

🚀 Setup Guide

1. Clone this repository
2. Install the required packages:

   pip install -r requirements.txt

3. Create a .env file in the root directory with your OpenAI API key:

   OPENAI_API_KEY=your_api_key_here
   

4. Run the Jupyter notebooks in order:
   • L1_Router_Engine.ipynb
   • L2_Tool_Calling.ipynb
   • L3_Building_an_Agent_Reasoning_Loop.ipynb
   • L4_Building_a_Multi-Document_Agent.ipynb

📖 Lesson Overview

Lesson 1: Router Engine

Learn how to create a router query engine that can direct queries to either a summary index or a vector index based on the query content.

Lesson 2: Tool Calling

Explore how to define and use tools with LLMs, including simple function tools and auto-retrieval tools for querying documents.

Lesson 3: Building an Agent Reasoning Loop

Build an agent that can reason over documents, maintain a conversation, and provide detailed responses with source attribution.

Lesson 4: Building a Multi-Document Agent

Create a sophisticated agent that can work with multiple documents, dynamically select the most relevant tools, and compare information across different sources.

🔗 Resources

• LlamaIndex Documentation
• OpenAI API Documentation

📝 License

This project is open source and available for educational purposes.