This paper aims to examine the implementation of LibraryGPT, a conversational artificial intelligence (AI) system integrated with the Tawk live-chat platform, in an academic library context. The authors describe a framework that embeds a GPT-based chatbot into the Royal Global University library’s Koha OPAC, outlining data collection and analysis of one year (May 2024–May 2025) of usage. Metrics include total interactions, chat durations and user satisfaction.
In this study, the authors recorded 216 user-chat sessions, with peak usage coinciding with key academic periods. The AI chatbot handled reference queries and provided personalized responses, yielding largely neutral user ratings. While most interactions were brief, occasional extended dialogues (over 20 min) demonstrated the system’s ability to assist with complex queries.
These results confirm that integrating AI tools in libraries can improve information retrieval and user engagement, in line with prior analyses of AI in libraries. The authors discuss how LibraryGPT supplements traditional reference services without replacing librarians.
The findings suggest that academic libraries can leverage AI chatbots like LibraryGPT and Tawk to expand services, but must address data quality, user training and ethical considerations. Future research should investigate long-term impacts on workflow and user learning.
