Site Reliability Engineer
You own production health: the single point of accountability for whether the system is up, fast, and secure.
Platform. Remote. Full-time. 4+ years. Hiring now.
You own production health: the single point of accountability for "is the system up, fast, and secure." Product engineers build, run, and secure their own services, so there is no separate DevSecOps role here. You own the platform underneath them, the reliability of the whole, and the cross-cutting security and ops work that isn't any one team's service.
What makes it different
You are not the first responder. An AI agent is. It triages every alert, pulls logs and metrics, finds the root cause, and arrives with a proposed fix. You approve, reject, correct it, and build the system that makes the next incident quieter. If you want to type kubectl from memory at 3am, this is not it. If you want to design and supervise the thing that stops most 3am pages from happening, it is.
What you own
- Production reliability: uptime, latency, capacity, and the incident response that protects them.
- The platform and cluster: EKS, data stores, Kafka, and the observability stack.
- Paging, escalation, and alert hygiene, which is the biggest and cheapest reduction in out-of-hours pages.
- Cross-cutting security and ops: cluster security posture, CSPM, IaC reconciliation, GitOps, and the security backlog.
- The AI first-responder: deploy it, prove its diagnosis accuracy, and raise how much it handles on its own.
DevSecOps is distributed, not a role
Each product engineer secures and operates what they ship. You set the standards, the tooling, and the guardrails, and you handle the platform-level and cross-service security work. This is desk work done during normal hours when people are at their desks, never mixed into firefighting an incident.
On-call and pay
Primary on-call is built into this role, and it is built into the salary. The 24/7 ownership is priced into base pay, not tracked by the hour and not paid per incident. There are no fixed working hours: you own an outcome, not a shift. Recovery after a night incident is expected (rest, not pay), planned time off is covered by a backup, and a sustained heavy load is a signal to review pay, add time off, or hire, not to grind. The job is to make pages rare.
This is not a rotation you join on top of another job. On-call is the job.
To apply, email us your resume.