Sr. Site Reliability Engineer

You will own production reliability across the platform — multi-tenant SLOs, observability, automation pipelines, incident management, secrets operations, and the infrastructure that keeps the platform running reliably for many clients simultaneously, each with their own SLOs and operational expectations.

You treat reliability as an engineering problem: if it can be measured, alert on it; if it happens more than once, automate it.

• Design and own the platform's reliability posture — define SLO contracts per tenant and service, set error budgets, and build alerting that pages on symptoms not causes. Make reliability a first-class engineering concern across every team.
• Eliminate toil through automation — build and maintain automation pipelines across the full operational surface: infrastructure provisioning, deployment workflows (blue/green, canary, rollback), self-healing runbooks, and auto-remediation for known failure modes.
• Own the observability strategy end-to-end — decide what gets instrumented, how signals are correlated, and how engineers get from alert to root cause fast. This includes AI-specific signals such as retrieval quality, model latency, agent success rates, and embedding drift — not just standard infrastructure metrics.
• Build and enforce the quality gate on every release — code changes run evaluation suites; regressions in model quality, retrieval accuracy, or agent task success block deployment. Own the pipeline that makes quality measurable and enforceable.
• Lead incident management — on-call, incident command, blameless postmortems with action items that actually close, and tenant-aware communication so each client gets the right information within their contractually agreed window.
• Own infrastructure as code across all environments — every environment is version-controlled, reproducible from scratch, and promotable through staging to production. Drift is detected and corrected automatically.
• Secure the operational surface — manage secrets lifecycle (rotation, distribution, customermanaged key support), maintain audit log infrastructure that is retained and immutable, and own supply-chain hygiene (dependency vulnerabilities, signed artifacts, software bill of materials).
• 7+ years of SRE, DevOps, or Platform Reliability experience, with 2+ years on multi-tenant SaaS in production.
• Strong on Kubernetes, containers, and infrastructure as code (Terraform, Pulumi, or CDK).
• Has built and maintained CI / CD pipelines at scale — GitHub Actions or comparable, branch protection, required checks, and deployment automation.
• Observability tooling: OpenTelemetry pipeline with at least one OTLP-compatible backend (Phoenix, Langfuse, Datadog, Honeycomb, Prometheus, or comparable).
• Incident management on a production system with paying customers — has been the person responsible when it goes down.
• Has automated away a meaningful piece of operational toil — can describe what it was, how they automated it, and what the measurable impact was.
• Infrastructure cost discipline — has managed and optimized a real cloud bill, not just observed it.