AI-Assisted Collection Management: The Gift That Keeps on Giving

The integration of artificial intelligence (AI) into cataloging and metadata workflows has gained increasing attention in library and information science. This study examines how advanced AI technologies, particularly retrieval-augmented generation (RAG), explainable AI (XAI), and AI-augmented optical character recognition (OCR), can support collection management services, using gift book processing as a testbed. The authors develop a RAG-enhanced workflow that processes book cover images, searches holdings across multiple library catalogs, and then assists in locating metadata for copy cataloging. The study also incorporates XAI and a human-in-the-loop approach to improve transparency, support iterative prompt refinement, and guide decision-making in workflow design, while offering practical guidelines for designing AI-assisted collection management workflows. The findings suggest that AI-augmented workflows, when combined with human oversight, can improve collection management practices and provide a practical workflow adaptable to libraries across a range of material formats.