We generate full-stack apps and deploy them — a generated app becomes a live URL, on our infrastructure or in the customer's own AWS or GCP account. It’s in production and it has more shell script in it than it should. You’d own it, and the mandate is to make it boring. Kubernetes, GCP and AWS deploy paths, CI/CD, observability, cost, tenant boundary. No on-call rotation today — it’s the founders and whoever last touched it. Building a real one is part of the job, not something you inherit alone. 4+ years production infra, Kubernetes past the tutorial, AWS and/or GCP at depth. The rest is on the page, including the open problems that don’t flatter us. I’m a co-founder — ask me anything here.