Senior DevOps Engineer, AI & Applications

<p><strong>Role Summary</strong></p> <p>Every AI feature we ship touches thousands of GPUs. The Senior DevOps Engineer will build the release engineering backbone—CI/CD pipelines, automated testing gates, one-click deployments with instant rollback—that lets Firmus scale fast and responsibly.&nbsp;</p> <p>You're the bridge between engineering and operations: setting Firmus standards for how code gets to production, mentoring the team on deployment safety, and driving a blameless culture when things go wrong. Ship safely. Ship often. Ship at scale.</p> <p><br><strong>Key Responsibilities</strong></p> <ul> <li>Design and maintain team-wide CI/CD pipelines (Jenkins, GitHub Actions, ArgoCD, or equivalent) with automated testing gates, artifact management, and deployments aligned with GPU cluster standards.</li> <li>Implement release engineering best practices: repeatable releases, GitOps workflows, automated rollback, and change management procedure.</li> <li>Build and manage test infrastructure: environment provisioning, data seeding, long-running job validation (especially for distributed training templates and multi-node job submissions).</li> <li>Establish engineering protocols and standards: repo organization, PR templates, code quality gates, dependency scanning, static analysis.</li> <li>Partner with infra teams to ensure AI product features deployment practices meet compliance and security standards for massive GPU clusters.</li> <li>Mentor team on testing strategies, deployment safety, and incident response procedures.</li> </ul> <p><br><strong>Skills &amp; Experience</strong></p> <ul> <li>5–7 years of CI/CD engineering, release engineering, or DevOps experience</li> <li>Deep expertise in GitHub Actions, GitLab CI, ArgoCD, or Jenkins with multi-stage pipeline design and testing gate implementation.</li> <li>Strong automation scripting (Python, Go, or Bash) for build orchestration and environment templating.</li> <li>Strong Kubernetes fundamentals (hands-on): deep understanding of Pod lifecycle and failure modes (Pending/Running/CrashLoopBackOff/Evicted), Deployments/ReplicaSets, Jobs/CronJobs, Services/Ingress, and how these primitives behave under load and during rollouts.</li> <li>Config &amp; secret management: practical experience designing and operating ConfigMaps and Secrets (including secret rotation patterns), with strong hygiene around least privilege, auditability, and preventing credential leakage into logs/artifacts.</li> <li>Safe rollout patterns: proven experience implementing and operating safe rollout strategies (rolling updates, canary, blue/green), readiness/liveness/startup probes, PodDisruptionBudgets, and rollback procedures—ensuring zero/low-downtime deployments for customer-facing services.</li> <li>Deployment safety &amp; debugging: ability to debug common Kubernetes rollout issues end-to-end (bad probes, misconfigured resources/limits, image pull failures, secret/config drift, node pressure/evictions) and convert learnings into automated CI/CD gates and runbooks.</li> <li>Familiarity with artifact management, versioning strategies, and rollback procedures.</li> <li>Experience integrating testing frameworks into CI pipelines (unit, integration, end-to-end).</li> </ul> <p><br><strong>Key Competencies</strong></p> <ul> <li>Engineering Velocity &amp; Time-to-Release improves quarter-over-quarter while release standards remain consistent (gates, tests, approvals, auditability).</li> <li>Platform Reliability &amp; Customer Trust remains strong: release-related incidents are rare and recovery is fast; reliability targets are met without "surprise outages."</li> <li>Developer Productivity &amp; Team Scale improves: engineers spend less time fighting CI/CD and more time shipping as the team grows.</li> <li>Cost Efficiency &amp; Resource Optimization improves: CI/CD and test infrastructure costs stay controlled (or decrease per unit of output) as usage scales.</li> <li>Knowledge &amp; Culture Multiplier effect is visible: release/reliability practices become the default across the org and repeat incident classes reduce</li> </ul> <p><br><strong>Success Metrics</strong></p> <ul> <li>Engineering Velocity &amp; Time-to-Release improves quarter-over-quarter while release standards remain consistent (gates, tests, approvals, auditability).</li> <li>Platform Reliability &amp; Customer Trust remains strong: release-related incidents are rare and recovery is fast; reliability targets are met without “surprise outages.”</li> <li>Developer Productivity &amp; Team Scale improves: engineers spend less time fighting CI/CD and more time shipping as the team grows.</li> <li>Cost Efficiency &amp; Resource Optimization improves: CI/CD and test infrastructure costs stay controlled (or decrease per unit of output) as usage scales.</li> <li>Knowledge &amp; Culture Multiplier effect is visible: release/reliability practices become the default across the org and repeat incident classes reduce</li> </ul> <p><br><strong>Location &amp; Reporting</strong></p> <ul> <li>Singapore or Australia (Launceston, TAS or Sydney, NSW)</li> <li>Reporting to Head of AI &amp; Applications</li> </ul> <p> <br><strong>Employment Basis</strong></p> <p>Full-time</p> <p><br><strong>Diversity</strong></p> <p>At Firmus, we are committed to building a diverse and inclusive workplace. We encourage applications from candidates of all backgrounds who are passionate about creating a more sustainable future through innovative engineering solutions.&nbsp;</p> <p>Join us in our mission to revolutionize the AI industry through sustainable practices and cutting-edge engineering. Apply now to be part of shaping the future of sustainable AI infrastructure.&nbsp;</p>