tonykipkemboi/ollama_pdf_rag

Implements a LangChain + ChromaDB vector pipeline with Ollama for embeddings and inference, eliminating cloud dependencies entirely. Offers three distinct interfaces—Next.js with REST API, Streamlit, and Jupyter notebooks—plus multi-PDF support with source citation tracking and multi-query retrieval strategies. Architecture combines FastAPI backend for document ingestion and RAG queries with a modern React frontend, enabling both programmatic and interactive exploration of document collections.

496 stars.