kreetikaa/rag-chatbot

DocMind is a small RAG (retrieval-augmented generation) document chatbot for uploading PDFs/TXT and asking natural-language questions about their contents. It returns answers with source citations and maintains conversational context. Intended for developers, analysts, researchers or anyone building/testing document-QA demos or prototypes. Key features: PDF/TXT ingestion with recursive chunking, embeddings persisted in ChromaDB, top-k vector retrieval, a LangChain RAG chain with a custom prompt, conversation memory, FastAPI backend (/upload, /query) and a simple frontend (Streamlit/HTML). Uses OpenAI GPT-4o-mini. Repo includes backend modules (ingestor, embedder, retriever, chain), frontend, and instructions: set OPENAI_API_KEY, pip install requirements, run uvicorn and open the frontend.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.