DevOps Engineer

Northwind runs a multi-region GPU fleet with hard latency commitments to customers. We're looking for an infrastructure engineer to own the reliability of that platform. This is not a ticket queue. You'll set the direction for how we do capacity planning, incident response, and cost control, and you'll have the autonomy to change things that aren't working.

• Own our Kubernetes platform across three cloud regions
• Build the observability stack — we use Prometheus, Grafana, and OpenTelemetry
• Drive down infrastructure cost per inference without compromising latency SLOs
• Lead incident response and run blameless postmortems
• Harden our deployment pipeline: progressive rollout, automated rollback, policy enforcement

• 5+ years in SRE, platform, or DevOps roles running production Kubernetes
• Strong Terraform or Pulumi — infrastructure is code here, with no exceptions
• Experience with GPU workloads or other cost-sensitive compute
• Track record of improving reliability metrics you can actually name

• Equity in a Series B company with meaningful upside
• Full medical, dental, and vision covered at 100% for you and dependents
• $2,500 annual learning budget, no approval required
• Four weeks paid vacation plus company-wide winter shutdown
• Home office stipend and a new laptop of your choosing

Northwind Labs builds the deployment and observability layer that AI teams use to ship models into production. Series B, backed by Amplify and Corner Ventures, with customers across developer tooling and fintech.