Authors: Dr. K. Himabindu, Faruk Ahmed, Pankaj Kumar Bara, Krishna Kumar Yadav, Soham Parmar
Abstract: Healthcare chatbots are gradually moving beyond simple question-and-answer systems by bringing together large language models, information retrieval, multimodal analysis, and human oversight. This paper explores the open-source Multi-Agent Medical Assistant as a safety-focused tool for providing medical information. The system includes separate components for input screening, conversation, retrieval-augmented generation (RAG), web-based evidence retrieval, medical image processing, response generation, and human validation. Rather than attempting to provide autonomous diagnoses, the system emphasizes evidence-based responses, transparent communication of uncertainty, and human review.
