We're looking for a Release Engineer (SRE) to join our Release Engineering team (part of EngOps) — a production-operations expert who brings an SRE mindset to how Supabase ships and runs, making deploys safe, observable, and recoverable at scale
Release Engineering's scope has grown well beyond build-and-ship: we increasingly own the operational reliability of the systems that deploy and run Supabase. In this role you'll treat our deployment pipelines, pre-production signal, and the control plane itself as production systems — with SLOs, error budgets, and on-call ownership — and you'll be the person teams lean on when reliability is on the line
This is not a "gatekeeper" role. You'll make the reliable path the easy path: standardising how we deploy, instrumenting what we ship, and ensuring that when something breaks, we detect it quickly and recover quickly
WHAT YOU'LL BE RESPONSIBLE FOR
Own the reliability of Supabase's deployment and release systems, and the control plane they run on, against clear SLOs and error budgets
Turn pre-production into a trustworthy signal — standardizing and instrumenting today's fragmented, ad-hoc deployment workflows
Drive disaster-recovery readiness, including making environments reproducibly deployable from scratch (untangling undocumented secrets, unclear configuration ownership, and circular service dependencies)
Build and operate health and SLO monitoring for critical user flows, using synthetic testing to catch regressions before customers do
Reduce mean-time-to-detect and mean-time-to-recover for deploy-related incidents — which account for a large share of our incident load
Participate in on-call, lead blameless postmortems, and turn findings into runbooks, alerting, and automation that remove toil
Improve deployment observability and auditability — a clear record of what shipped where, when, and by whom
Document operational procedures — break-glass paths, access models, and runbooks — so reliability knowledge isn't tribal
Define and track SLAs, SLOs, error budgets, and DORA delivery metrics — with meaningful alerting over noise
Ensure deployments fail fast and safely when health checks degrade
Harden access and break-glass workflows (e.g. scoped self-service) so the right people can act in an incident without unsafe workarounds
Partner with product engineering and platform teams to align release practices with reliability and availability targets