AyurSanvaad: A Multilingual Chatbot for Ayurvedic consultation using LLMs

DOI

10.2991/978-94-6463-831-8_41How to use a DOI?

Keywords

Ayurveda; Retrieval-Augmented Generation (RAG); Large Language Models (LLMs)

Abstract

The convergence of traditional healthcare and modern AI offers a unique opportunity to improve access to ancient systems like Ayurveda. We present AyurSanvaad, a multilingual RAG-based chatbot that delivers contextually accurate and semantically faithful Ayurvedic responses. It integrates a lightweight language model with multilingual detection, translation, and dual semantic search backends, enabling natural interaction in over 22 Indian languages. User queries are auto-detected, translated to English, retrieved from a curated Ayurvedic knowledge base, and retranslated into the user’s language. RAGAS evaluation shows AyurSanvaad achieves up to 13% higher faithfulness, with notable improvements in answer and contextual relevance.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Saket Pol
AU  - Atharv Malusare
AU  - Tejas Parchure
AU  - Jeet Narkhede
AU  - Swapna Yenishetti
AU  - Lakshmi Panat
PY  - 2025
DA  - 2025/08/31
TI  - AyurSanvaad: A Multilingual Chatbot for Ayurvedic consultation using LLMs
BT  - Proceeding of the 1st International Conference on Lifespan Innovation (ICLI 2025)
PB  - Atlantis Press
SP  - 339
EP  - 346
SN  - 2468-5739
UR  - https://doi.org/10.2991/978-94-6463-831-8_41
DO  - 10.2991/978-94-6463-831-8_41
ID  - Pol2025
ER  -