Risky infrastructure, made boring.

I work across the systems a product depends on, wherever a routine change can turn into an incident. When it’s built right, a risky change is controlled instead of stressful, and a bad day shorter than it would have been.

Simple by default: managed services and proven primitives before the clever option, a bespoke tool only when the problem actually needs one, scoped tight to the gap.

The hardest part isn’t the tooling. It’s saying no to the next one-off hack that someone will swear is just for now.