Back to projects
PythonLangChainChromaDBOllamaFastAPIDocker

RAG Document Query System

An experimental RAG system that indexes technical documentation and codebases into a vector database, enabling semantic search and natural language querying. Uses local embedding models for privacy and cost efficiency.

February 10, 2025
Source code

Key Highlights

  • Document chunking with semantic boundary detection
  • Local embedding generation using Ollama models
  • ChromaDB vector store with similarity search
  • FastAPI endpoint with streaming response support
  • CLI tool for document ingestion and indexing