Medical Document Q&A Assistant (RAG)
Built a RAG-based medical document Q&A assistant to enable question answering over clinical documents with citations and streamed responses. Implemented hybrid retrieval to improve relevance across the underlying document corpus. Integrated a React frontend to support interactive querying and answer streaming for end users. • Built LangChain + ChromaDB RAG system for natural language queries. • Implemented hybrid retrieval and source citation. • Added streaming responses via React frontend. • Worked with clinical document text as the primary content type.