Lead DevOps Engineer

We're looking for an ambitious, hands-on Lead DevOps Engineer to build and scale the reliability and infrastructure function at […]. Our platform runs AI coding agents in cloud sandboxes, many times in parallel, to build and quality-check reinforcement-learning tasks for frontier AI labs. Every run is real, paid compute.

The platform ships to production continuously, and it has to stay working and available while it is being built. This is an opportunity to own that from zero to one, establish the practices from scratch, and potentially grow into leading the team that keeps it running. What You'll Do Build reliability from 0 to 1: Establish on-call, incident response, SLOs and runbooks where none exist yet, and keep the platform available while it changes underneath with heave AI coding.

Scale the platform horizontally: Move process-local coordination into shared state, so our workers can run as many replicas without double-booking capacity or paying for work twice. Run the data layer: Own PostgreSQL, Redis and the job queue in production, including connection budgets, pooling, migration safety, backups and restore drills. Make failures loud: Build observability with Prometheus, Alertmanager and Grafana that ends at a person who can act, not at a dashboard.

Own the DevOps foundation: CI/CD in GitHub Actions, infrastructure as code with Terraform on GCP and Render, reproducible container images, and secrets management. Treat cost as an engineering metric: Find the unit dr