Agentic RAG for Dummies

Active
GitHub Jupyter Notebook MIT

Description

An educational Agentic RAG project with clean code demonstrating how to build RAG systems with agent capabilities — routing, retrieval, evaluation, and iterative refinement.

Key Features

  • Hierarchical indexing with parent/child chunks for precision and context
  • Conversation memory for multi-turn dialogue continuity
  • Query clarification with rewrite and human-in-the-loop pause
  • Multi-agent map-reduce for parallel sub-query processing
  • Self-correction with automatic re-query on insufficient results
  • Langfuse observability for LLM calls, tool usage, and graph execution

Use Cases

💡 Building modular RAG systems with agent-driven retrieval
💡 Learning agentic RAG concepts through interactive notebooks
💡 Processing multi-part questions with parallel sub-agents
💡 Creating context-aware conversational AI over document collections
💡 Swapping LLM providers (Ollama, OpenAI, Anthropic, Google) in one line

Tags

rag agent tutorial python llm

Categories

📚 RAG Tools

Quick Start

Open the Colab notebook for a guided walkthrough, or clone the repo, install dependencies, configure your LLM provider (Ollama/OpenAI/Anthropic/Google), and run the modular project.
Visit GitHub

Related Projects