RAG-based Chat Application using LLMs: A Case Study of Vikram Sarabhai Library IIM Ahmedabad

Abstract

As a customer-centric organization, the library constantly focuses on providing the best user experiences with respect to accessing quality information in the shortest possible time. In the current digital world, the library needs to have communication and support methods that are both efficient and real-time to ensure a high level of user satisfaction and engagement. The Vikram Sarabhai Library (VSL) designed a pilot project to develop a cutting-edge Retrieval Augmented Generation (RAG) based chat application using Large Language Model (LLM) technology. This paper presents an abstract of the project, highlighting its objectives, methodology, key features, and anticipated impact. The primary objective of the pilot project is to revolutionize user interaction and support services by harnessing the power of RAG-based chat capabilities. By combining advanced Natural Language Processing (NLP) techniques with Gen AI capabilities, the application aims to provide personalized and accurate responses to user queries in real-time from Authentic sources. Key features of the chat application include retrieval-based responses for factual queries, generation-based responses for nuanced inquiries, personalized recommendations based on user preferences, and seamless integration with VSL's existing knowledge base. By leveraging RAG-based techniques and custom datasets, the chat application aims to significantly enhance user satisfaction, streamline information retrieval processes, and foster a more engaging and interactive library experience for patrons. The Retrieval-Augmented Generation (RAG) based custom datasets chat application using LLMs is a testament to VSL's commitment to innovation and excellence in user-centric services. Through this pilot project, VSL sets new benchmarks in digital communication, support, and knowledge dissemination within the library community.