The adoption of OPACs from the 1980s onwards prompted large-scale retrospective conversion efforts across research libraries, but that work was never universal. Smaller or specialised collections, subject catalogues, and other complex card sequences were frequently left behind — their contents still dormant in card drawers or locked inside scanned images.

Large (vision) language models offer a new set of tools for this problem, and a number of library teams have begun working with them on catalogue cards. This workshop brings together practitioners from research libraries currently engaged in this work to share experiences and compare approaches. Rather than showcasing finished solutions, it is oriented around open exchange — invited projects presenting their work, followed by discussion open to all attendees.