An In-House Approach to AI: Developing a Custom Chatbot for Library Services

Abstract This article presents the University of Manitoba Libraries’ (UML) initiative to develop an in-house AI chatbot, a project aimed at enhancing library services through innovative technology tailored specifically to user needs. The chatbot was created to complement existing reference services by triaging routine inquiries and escalating complex questions to human librarians, ensuring personalized support. The article details the project’s stages, including initial research, requirements gathering, design and development, and iterative testing. This case study offers valuable insights and a model for other academic libraries considering similar AI-driven service enhancements.