Fit to programme

This was a proposed solution answering Task 027: Exploration of AI-assisted operations, behind WP 2.3.

Summary

We propose an Agentic Observability Framework: a self-hosted solution designed to transform the HPC experience for both users and staff. By transitioning from passive monitoring to active, AI-driven diagnostic support, this project addresses the growing operational complexity inherent in modern HPC environments. This framework enhances user self-sufficiency through intelligent assistance while enabling Research Technology Professionals (RTPs) to provide more rapid, high-quality advice. By automating routine triage, we ensure that expert human resources—a finite and vital asset—are optimised for the most intellectually challenging and demanding tasks, rather than being diverted by repetitive support requests.