Zero-hallucination AI Retrieval
shippedDocQA — RAG Document Q&A Engine
Full RAG document Q&A system built with FastAPI, sentence-transformers local embeddings (zero API dependency), ChromaDB vector store, and Groq Llama 3.3 for grounded answers with strict source citation.
- PDF Chunking
- Sentence Embeddings
- ChromaDB Vector Match
- Re-ranking Filter
- Llama 3.3 Answer
- Python
- FastAPI
- ChromaDB
- Llama 3.3
- Docker
- React
100%grounded response rate with citations