How UTS developed a suite of Course Handbook Bots

In December 2023, UTS convened a generative AI workshop to develop its first proof-of-concept applications. Supported by Microsoft, several teams proposed and tested concepts, one being the use of Large Language Models (LLMs) and conversational agents to revise the content of the UTS Course Handbook. This was to help the Education Portfolio accelerate the migration and modification of content to a new curriculum platform, incorporating human academic input throughout. The task of redrafting descriptions of Courses and Subjects to a consistent length, in a more student-friendly tone, seemed, in principle, ideally suited for LLMs. In addition, we sought to provide about six Course Intended Learning Objectives (including an Indigenous CILO) and Subject Learning Objectives for all handbook entries, which required distilling the existing CILO/SLO lists in a way that was acceptable to academics.

This formidable, pan-university challenge needed to be completed on an extremely tight timescale, with a lot of organisational learning en route. In this webinar, you’ll hear first-hand from the diverse members of the team on their experiences with generative AI, designing and refining a suite of custom chatbots, and from faculty academics on their experience receiving the revised content.

Speakers

• Simon Buckingham Shum (Connected Intelligence Centre)
• Susan Gibson & Kathryn Fogarty (Data Analytics & Insights Unit)
• Sharon Coutts, Josh Dymock & Danny Tan (Teaching, Learning & Curriculum Unit)
• Antonette Shibani (Transdisciplinary School)
• Sally McCarthy (Indigenous Teaching and Learning Team )
• Amanda Lizier (Faculty of Arts & Social Sciences)
• Lee Wallace (Faculty of Health)