Why my CI got 4 minutes faster

Latency budgets work better than latency targets. Targets are aspirational. Budgets force you to delete features when you go over.

Caches solve a problem and create three new ones: invalidation, observability, and the slow drift between what the cache thinks is true and what the source thinks is true.

I trust monorepos more in the second year than I did in the first. Most of the early friction goes away once the tooling catches up.

Most retry logic I've removed turned out to be hiding a real bug. The retries made the failure invisible, and the bug compounded for months in the background.