Overview

Isambard-AI provides an unprecedented AI research resource for the UK. Attracting thousands of researchers from academia, government and industry covering every conceivable research domain, users of the Isambard-AI do not always have a traditional HPC background. This diverse and innovative user community has exhibited unprecedented user patterns, heavily leveraging AI Agents and LLMs. This has presented novel challenges both in terms of technical guardrails and user guidance.

In this case study you will discover the progress Isambard-AI has made in developing 'Agentic Ops' to help protect the service and maximise user productivity.

This case study was written by Dr. Richard Gilham, AI Supercomputing Infrastructure Specialist, Bristol Centre for Supercomputing, University of Bristol. Richard has been at the Bristol Centre for Supercomputing since 2024. His main focus is leading the team’s work on user facing services, aimed at maximising user productivity and minimising friction. Prior to this he was the manager of the UM System Team at the Met Office.

Who is this case study for?

Teams who have noticed a change in user patterns from AI Agents and LLM's on their services.

Case Study Details

The slides are available here or you can watch the recording below.