3 posts
I'm building what might be the world's largest internal deployment system. In 2025, we expanded it from internal bare-metal infrastructure to native cloud resources. Here's what happens when you need to deploy to dynamic fleets defined by tags and integrate with autoscaling.
When automation systems need to know if they can touch a resource, 'is this production?' becomes a surprisingly complex question. I built a service to classify the productionness of infrastructure—it's harder than you think.
How do you deploy security patches across thousands of compute instances spanning multiple regions without causing an outage? Turns out, infrastructure-as-code starts to break at scale. Here's what I learned building a global patching system.