Why this role exists :

Razorpay moves money for millions of businesses. Hundreds of services ship to production through a delivery platform we built on Spinnaker across multiple geographical regions. That path from commit to production is our biggest lever on both production risk and engineering speed: Several outages arrive through it, and every team's shipping speed is capped by it.

You will own that path end to end, from build to deploy, as a product with internal customers.

You'll lead change across product teams through influence, not authority. You'll make the paved path the default, enforce standards at the pipeline gate, own migrations yourself with tooling rather than tickets, and make the safe path the faster one. You'll use leadership support for deprecations when the data justifies it.

What you'll own :

- Build and CI: a standard way for services to build and ship artifacts, with traceability from commit to production, build speed treated as a tracked goal.
- Continuous delivery: progressive delivery as the default (canary with automated analysis and auto-rollback, multi-day bake), safe coordinated releases across regions, and proving pipeline changes before they reach production.
- Safety and compliance: deployment rules based on service criticality, enforced in the pipeline; policy as code; and change records that satisfy PCI-DSS and RBI requirements automatically.
- Adoption: the share of services on the paved path, not just the fact that it exists.

You will succeed with :

Deep expertise in production deployment systems: orchestration, rollout strategy, blast radius,



rollback and multi-region coordination. Strong working competence in build and CI infrastructure, which we expect you to grow into. If you're deep in both, tell us; it changes the level we hire at.

What we're looking for :

- 8+ years building and operating production systems, with Staff-level scope: org-wide technical direction and at least one cross-team migration driven to completion.
- You've driven a technical decision through teams that didn't report to you, against real resistance. This is the most key requirement.
- You've built an internal platform other teams adopted, and can explain why they adopted it.
- Deep experience designing and running CD at scale (Spinnaker, Argo CD/Rollouts, Harness or an in-house equivalent), including its failure modes.
- Deep Kubernetes knowledge: rollouts, apply and patch behaviour, and failure handling.
- Strong backend engineering in Go, Java/Kotlin or Python, and comfort working in large codebases you didn't write, including open-source tools.
- Sound judgement on the trade-off between speed and safety, defensible with data to engineers and executives.

Nice to have :

Contributions to Spinnaker or similar open-source projects; statistical canary analysis; workflow engines (Temporal); hermetic builds and remote caching (Bazel); supply-chain security (SLSA, Sigstore); and delivery under audit requirements in finance or another regulated industry.

What this role is not :

Not a release manager, not a deploy-button or ticket-queue role, and not just "keep CI running." You'll be expected to change the architecture of both CI and CD.

📌 Principal Engineer, Release Engineering (Bengaluru)
🏢 Razorpay
📍 Bengaluru