A flexible infrastructure for a teaching-led AI-empowered university

This project evaluates the optimum infrastructure for academic staff at UCL to develop LLM-based teaching tools.

This project explores the infrastructure and pedagogical potential of integrating large language models (LLMs) into UCL’s teaching practices. By developing two prototype tools—one in teacher education and one in medicine—the project investigates how AI can support knowledge transfer and semi-automate assessment and feedback processes.

The initiative evaluates student engagement with these tools and identifies the technical and institutional infrastructure needed to scale their use. Findings will be shared across UCL to inform future practice and policy, supporting the university’s ambition to become a data-empowered, AI-enabled institution.

By focusing on high-impact domains, the project demonstrates how AI can enhance teaching quality, reduce workload, and foster innovation. It also lays the groundwork for a sustainable model of AI integration that empowers academic staff to co-create learning tools tailored to their disciplines.