About
I'm a staff software engineer. I spent nearly nine years at Meta, seven of them owning performance and reliability for Instagram on Android.
The work I'm best at starts with a symptom nobody can explain: a crash rate that won't move, a screen that judders, an incident that keeps coming back. I find the cause, fix it, and then build whatever stops it returning, whether that's a lint rule, a dashboard, a training programme or an agent that does the first hour of debugging.
What I care about
Craft. I write down the options I rejected as well as the one I chose, and this site holds itself to a published performance and accessibility budget.
Leverage through people. The Reliability Captain programme I founded trained 12 engineers to run incidents, and it kept running after I moved on.
Open work. This site has no tracking scripts, and its source is public.
Safe, useful AI. I build agents that debug production systems, and I write about where they fail as well as where they help.
What's next
I'm looking for work where a large system needs to get faster, steadier or easier to operate. That covers SRE, performance, platform, Android and AI infrastructure.