The Library Association of Singapore (LAS) Training and Development Committee launched its inaugural Lunchtime Brown Bag Series on 17 September 2025, with 63 participants in attendance. This new initiative aims to facilitate knowledge sharing and professional discussions amongst library professionals.

Phoebe Lim, Chair of LAS Training and Development Committee, opened the session by outlining the series’ objectives to foster meaningful exchanges among librarians.

Presentation: NLB’s AI and Automation Experiments

Abigail Huang, Principal Librarian with the Resource Discovery team, and Jeremy Goh, Metadata Services Librarian, both from the National Library Board (NLB), presented on their organisation’s experimentation with automation and artificial intelligence for metadata creation. Their presentation covered NLB’s trials using AI to catalogue the WebArchiveSG collection and the development of “SGCAT,” a custom GPT prototype designed to catalogue physical library materials. The speakers discussed both the potential benefits and current limitations of these emerging technologies.

Presentation slides: LAS Brown Bag_20250917_AI and automation in metadata at NLB

Glossary and links for 17 Sep LAS Brownbag

Q&A Session Highlights

The presentation was followed by an engaging question-and-answer session, with participants raising several pertinent queries:

• The transition of human cataloguers from creators to reviewers in AI-assisted workflows
• Service provider recommendations and partnerships
• Guidance for organisations interested in beginning their own AI experimentation, including vendor versus in-house development considerations
• Integration of Linked Data with authority files
• Future plans for AI implementation in subject analysis, subject heading assignment, and call number generation

Breakout Room Discussions

Participants engaged in small group discussions centred around four key questions exploring the role of AI in library work. The discussions were framed by a thought-provoking tweet from @AuthorJMac questioning whether AI should handle routine tasks to free professionals for creative work, or vice versa.

One breakout room, comprising five members from INSEAD, NUS, Nanyang Polytechnic (NYP), and NLB, shared diverse perspectives on AI implementation across different institutional contexts. The NYP representative highlighted challenges with eBook metadata and expressed uncertainty about where to begin with AI integration in cataloguing workflows. The NUS participant described using existing AI chatbots to assist with name suggestions for materials outside their subject expertise. Meanwhile, the INSEAD representative detailed their work developing a generative AI agent to process research papers and extract cataloguing information such as DOIs and author details, though currently operating at a small scale.

The INSEAD participant also noted significant challenges including copyright restrictions that prevent the use of public AI tools, necessitating private, internal server solutions such as AWS. Their team is additionally working on building a knowledge graph to enhance AI functionality, though time constraints prevented further elaboration on this initiative.

Emerging Insights

The session highlighted the varied stages of AI adoption across Singapore’s library community, from initial exploration to pilot implementations. Common challenges identified included technical barriers, copyright considerations, scalability issues, and the need for guidance on implementation strategies. The discussions underscored both the potential of AI technologies to transform metadata workflows and the practical considerations that institutions must navigate in their adoption journey.

____________________________________________________________________________________________________________________________________________________________________

Discover more about LAS’s professional sharing and training programmes. Click here to find out more!