System Admin

Good infrastructure is invisible. Nobody notices uptime; they only notice outages. My goal on every system I run is to make it boring on purpose — calm enough that nobody needs a hero moment, and predictable enough that a 2am page is a genuine surprise rather than a Tuesday.

That means real monitoring instead of vibes, deployment processes that don't depend on someone remembering a manual step, and infrastructure-as-code so the setup is reproducible instead of tribal knowledge living in one person's head. I've run this across Linux boxes, containers, and the big three clouds — the tools change, the discipline doesn't.

Currently that discipline is keeping a multi-location web and growth stack at 99.9% uptime — not glamorous, but it's the kind of number that only happens when the boring stuff is actually done right.