RAGNet: Enhancing Conversational AI in Medical Domain
Nihir Varshney, Saksham Mantoo, Madhav Wadhwa, Mohammad Aatif, Shweta Bhardwaj · 2025
An Enhanced version of the RAG framework is proposed to improve the Groundedness of Medical Question Answering systems. It achieves higher retrieval effectiveness and response generation due to advanced techniques, such as few-shot prompting, multi-hop retrieval, and query optimization. In this study, the comparison models used are the Naive and Advanced RAG models for large language models on medQA dataset. Application in the field of practice would highlight this topic in medical applications where precision matters and the Enhanced Modular RAG provided one of the most reliable answers to challenging medical questions.