Skip to content

Delivery evidence

The course delivery must be documented on GitHub and reviewed by the group. This page is the index for the evidence that should accompany the final submission.

Required evidence

  • Contributors and service ownership.
  • Architecture diagram and repository map.
  • API contracts and examples.
  • Local setup and deployment runbook.
  • AWS account, region and resource inventory without secrets.
  • EKS topology and Kubernetes manifests or commands.
  • CI/CD pipeline runs and rollback procedure.
  • Load-test scenarios, raw results and interpretation.
  • Bottleneck investigation and before/after comparison.
  • Cost estimate, actual spend and PaaS decisions.
  • SLO/SLA definitions and observability dashboards.
  • Test strategy and final verification.
  • Known limitations and future work.

Staging backlog — not implemented

  • Implement and accept the developer-only staging plan: fixed small copies of all production business services, scheduled availability, synthetic seed data, Cloudflare Access, separate Grafana Cloud observability, Faro errors enabled and Umami/usage analytics disabled.
  • Record the approved budget/schedule, hosting/parity tradeoffs, isolation tests, independent telemetry gates, recovery/rollback and operating evidence.

These items are TODO; documenting the plan does not complete them.

AI disclosure

The course permits limited AI assistance for tasks such as boilerplate, improvements and error fixing. The delivery must state:

  • which artifacts or code were assisted by AI;
  • what prompts or task descriptions were used when relevant;
  • what the group reviewed, tested and changed;
  • which parts were implemented directly by group members.

Every group member must understand and be able to defend the complete project, including AI-assisted work.

Reproducibility

For each claim, link to a commit, pull request, workflow run, dashboard export, load-test report or command output. Include the date, environment and version. Never include credentials, private tokens or personal data in evidence.