Each participant builds and demos a mini Internal Developer Platform for a fictional healthcare or payments company. A new service goes from a portal click to production, then survives a seeded failure.
Infrastructure provisioned from Terraform modules through a plan-on-PR pipeline with LocalStack and remote state.
One-click service creation generating repository, CI pipeline, Argo CD application, SLOs, and dashboards automatically.
Application deployment via Argo CD with signed images (cosign) and Kyverno security admission policies enforced.
Telemetry pipeline capturing traces, metrics, and logs with automated burn-rate alerts generated via Sloth/OpenSLO.
Saturation testing using k6 and cost-per-service analysis using OpenCost.
Handling seeded production failures through alerts, rollbacks, or Velero restores, followed by a blameless postmortem.
Custom Kubernetes controller written in Python/Go that safely remediates known failure patterns with guardrails.
