MATERNAL HEALTH BOT USING RAG ARCHITECTURE

Abstract This study describes the deployment of a voice-activated chatbot designed to assist expectant mothers by offering trustworthy maternal health information. To provide precise and contextually aware responses, the system employs a Retrieval-Augmented Generation (RAG) technique, which combines a language model with a local knowledge store. We developed this totally with free and open-source technologies to make the solution more accessible in rural or low-resource environments. Open-source approaches, such as Whisper for speech-to-text, are used to add voice capabilities. To ensure accuracy and validity, the data is also gathered from reputable sources like the WHO and other health portals. We used PDFs from these sources for RAG and stored them in vector databases for efficient document retrieval, as this system aims to bridge the information gap. Keywords Maternal health, RAG, chatbot, voice interface, NLP, vector database