Version deltas get called out early—helpful when juggling prod upgrades across clients and you don’t want to diff docs for hours. In Module 4, the v1.12→v1.13 migration table plus the PR walkthrough where CI flakes at ~8k RPS stuck with me; seeing the exact config flip made it faster to patch my repo. The framing stays client-outcome focused: fewer regressions, quicker merges, and clearer arch decisions that play nice with infra and obs. Mostly works, though I wasn’t sold on the brief k8s autoscaling bit; a little more math would’ve saved a Slack thread. still, the pacing fits between meetings and I’ve already reused the checklist on a live rollout. The tradeoff section is the part I keep open in a tab for reference.