Scaling Self-Hosted Compute: A lesson in Platform Engineering across an enterprise

The session will cover a 1.5 year journey to finally create the one platform to rule them all! - powered by GitHub and GitHub Actions. The main lessons learned will cover onboarding and compute. How do we scale our self-hosted compute to be as extensible as possible across an org with demands for high-spec’d compute for AI workloads? Enter ARC (actions-runner-controller), GitHub’s solution that orchestrates and scales self-hosted runners for GitHub Actions. I will walk everyone through the journey on self-hosting compute at an enterprise level with lessons learned in caching, label definitions, and small tidbits from hosting VMs instead of K8 clusters.

Speaker

imran

Imran Ali


I’ve been in the industry for about 6 years now, currently contracting for CIBC where I build out the platform to enable teams to deploy their code to production quickly, safely, and

...