Document Q&A System (RAG)

This is an educational implementation of a RAG (Retrieval-Augmented Generation) system that exposes its inner workings. Upload documents and ask questions while observing the complete pipeline: document chunking, vector embeddings via ChromaDB, semantic search, and GPT-4 interactions.

The system logs every API call, vector lookup, and context injection, showing you exactly how data flows between components. Built with Python and Flask, packaged in Docker, and designed for developers who want to understand RAG systems at the implementation level.