AyUR-bot: A Revolutionary Ayurvedic Chatbot Empowered by Generative AI
Ratna Nitin Patil, Suyash Yeolekar, Omkar Surendra Khade, Yash Kadam, Prajwal Ingole, Suraj Patil · 2024
This study introduces AyUR-bot, a 24/7 Ayurvedic expert chatbot designed for prompt addressing of user queries on nutrition, remedies, and overall wellness. Project objectives include developing an Ayurvedic chatbot using NLP and deep learning, providing accurate responses, optimizing model efficiency, educating users about Ayurveda, and contributing to health and wellness through interactive conversations. The user-friendly chat interface allows effortless query submission. AyUR-bot leverages NLP tools, integrating the Llama 2 Large Language Model (LLM) and implementing Retrieval Augmented Generation (RAG) for question answering, optimizing Ayurvedic knowledge accessibility. The development process involves curating a specialized Ayurvedic document repository and utilizing the Llama 2 LLM model from Hugging Face. AyUR-bot ensures precise responses with unique stopping conditions. Components like the conversational retrieval chain, recursive character text splitter, and embeddings generator synergize for comprehensive responses. Evaluation with BERT Score validates AyUR-bot's effectiveness in delivering context-aware insights. The user-friendly interface enables interactive engagement, facilitating personalized Ayurvedic guidance. This research converges chatbot technology, NLP, and Ayurveda, aligned with project objectives, catering to the growing interest in traditional wellness, democratizing access to Ayurvedic wisdom. AyUR-bot results in a significant advancement in personalized Ayurvedic assistance.