RAG Based Medical Chatbot

Project Overview

A Retrieval-Augmented Generation (RAG) medical assistant chatbot designed to provide context-aware, reliable health information by retrieving relevant knowledge from curated medical literature and passing it to large language models.

Key Features

• Knowledge Retrieval Pipeline: Chunked and vectorized medical reference documents into high-dimensional vector embeddings using LangChain and HuggingFace models.
• Vector Database: Stored and indexed embeddings with Pinecone for fast, scalable nearest-neighbor semantic search.
• Context-Augmented Generation: Integrated Google Gemini API to formulate accurate, grounded answers based solely on retrieved domain context.
• Web Interface: Built an interactive, user-friendly chat interface using Flask and responsive frontend design.

Tech Stack

• LLM & Embeddings: Google Gemini API, HuggingFace Embeddings, LangChain
• Vector Store: Pinecone
• Backend & Web: Python, Flask, HTML/CSS, JavaScript