albertodiazdurana/rag-document-assistant
Production-ready RAG system with multi-provider LLM support (OpenAI, Claude, Ollama), vector database integration, FastAPI backend, and MLflow evaluation. Features German language support and Streamlit UI.
What's novel
Production-ready RAG system with multi-provider LLM support (OpenAI, Claude, Ollama), vector database integration, FastAPI backend, and MLflow evaluation. Features German language support and Streamlit UI.
Code Analysis
19 files read · 4 roundsA RAG document assistant that ingests PDF/MD/TXT files into ChromaDB, retrieves relevant chunks via embeddings, and generates answers using multi-provider LLMs (OpenAI/Anthropic/Ollama) through a FastAPI backend and Streamlit UI.
Strengths
Clean modular architecture with consistent patterns, well-documented code, multi-provider LLM support, and a thoughtful cross-lingual experiment with proper academic grounding. Tests are well-structured with meaningful mocks and edge cases.
Weaknesses
The core RAG chain is a thin LangChain wrapper with significant code duplication across invoke/stream/astream, conversation history is broken in the API (new chain per request), evaluation metrics are toy-level keyword overlap, and 'production-ready' is overstated (wide-open CORS, global mutable state).
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
22
Contributors
1
Files
59
Active weeks
3
Repository
Language
Python
Stars
2
Forks
0
License
—