Healthcare Chatbot: AI-Powered Medical Information Assistant Using Retrieval-Augmented Generation

Abstract

The rapid advancement of Artificial Intelligence and
Natural Language Processing has transformed modern
healthcare information systems. This paper presents a
Healthcare Chatbot developed using Retrieval-Augmented
Generation (RAG) architecture to provide accurate,
reliable, and context-aware healthcare responses. The
proposed system retrieves medical information from
trusted healthcare documents using semantic vector search
before generating responses through a Large Language
Model. Hugging Face embedding models are used to
generate semantic vector embeddings, while Pinecone
vector database stores and retrieves contextual healthcare
information efficiently. LangChain orchestrates retrieval
and prompt generation workflows, and the Groq API
provides access to the Llama 3.3 large language model for
high-speed response generation. The chatbot is
implemented using Flask and deployed using Docker and
AWS EC2 cloud infrastructure. Experimental results
confirm improved factual accuracy, reduced hallucination,
semantic understanding, and fast response generation
compared to traditional generative AI systems. The
proposed Healthcare Chatbot demonstrates how modern
AI technologies can improve healthcare information
accessibility for ordinary users through conversational
interfaces.