⚡ A modular Retrieval-Augmented Generation (RAG) stack with FastAPI backend, Matrix bot integration, and data ingestion pipeline using Qdrant.

⭐ If you like this project, don’t forget to star the repo!

• 📂 Data Ingestion → Upload & index documents into Qdrant
• 🤖 Matrix Bot → Chat with the RAG system inside Matrix rooms
• 🚀 FastAPI Backend → Serve embeddings, retrieval & generation endpoints
• 🔑 Session Management → Powered by Redis
• 🐳 Containerized with Docker Compose
• 📝 Built with LangChain, OpenAI, Docling, Qdrant

├── rag_backend/ # FastAPI app (RAG endpoint) │ ├── requirements.txt │ ├── Dockerfile │ └── tests/ │ ├── matrixbot/ # Matrix Nio bot for chat interface and Redis session │ ├── requirements.txt │ ├── Dockerfile │ └── tests/ │ ├── data_ingest/ # Ingest & index data into Qdrant │ ├── requirements.txt │ └── tests/ │ ├── docker-compose.yaml └── README.md 

git clone https://github.com/yourname/rag-project.git cd rag-project 

docker-compose up --build 

• FastAPI docs → http://localhost:8000/docs
• Qdrant UI → http://localhost:6333/dashboard
• Matrix Bot → Join your Matrix room

• Support multiple LLM providers (Anthropic, DeepSeek, etc.)

• Add PDF/Docx ingestion pipelines

• Extend Matrix bot with custom commands

PRs are welcome! Please follow conventional commits.