About The RoleAbout The Role:Soar runs a growing portfolio of financial-services and enterprise products on Google Cloud, with a disaster-recovery footprint on Alibaba Cloud. The platform spans around ninety Cloud Run services and several GKE clusters, deployed from more than a hundred repositories, provisioned with Terraform, and operated alongside an AI DevOps agent that proposes infrastructure changes for humans to approve.
We are hiring two senior engineers who are equally at home in DevOps, cloud architecture and SRE to own this platform end to end — deploy paths, infrastructure as code, Kubernetes, secrets and identity, reliability and disaster recovery — and to set the standard the rest of engineering follows.
Key Responsibilities:
- * Deploy paths: Converge our Cloud Run and GKE deployments onto one reviewed pattern — GitHub Actions with keyless Workload Identity Federation, additive configuration semantics, every change through a pull request — working repository by repository with the teams that own them.
- Infrastructure as code: Grow our Terraform estate (Google, Google-beta, MongoDB Atlas and Alibaba providers) from foundations — VPC, IAM, security, zero-trust access — through workloads to Kubernetes, with remote state, plan-in-PR, apply-on-merge and a drift policy people actually follow.
- Kubernetes: GKE operations end to end: node pools, Workload Identity, least-privilege RBAC, Helm charts, Istio / Gateway API ingress, backup and restore, zero-downtime upgrades.
- Secrets and identity: Drive the move from inline configuration to Secret Manager references; own service-account and IAM hygiene; make the audit trail the answer to "who changed what, when, approved by whom".
- Reliability: Define SLOs for the services that matter, build alerting and dashboards (Cloud Monitoring, OpenTelemetry), run incident response, write blameless postmortems and turn each one into an engineering change.
- Disaster recovery: Keep the Alibaba Cloud DR footprint — ACK, VPC, KMS, OSS — provisioned by Terraform and proven by regular exercise.
- The AI DevOps agent: Our agent proposes IAM grants, restarts, configuration and Cloud Run changes; humans approve; the platform executes and records. You will shape what it may propose, review the pull requests it opens and harden its guardrails. You are the human in that loop.
RequirementsRequirements:
- 6+ years building and running production infrastructure, 3+ on Google Cloud at depth: GKE, Cloud Run, IAM, Secret Manager, VPC and networking, Artifact Registry, Cloud Build, Cloud Logging and Monitoring.
- Terraform in anger module design, state management, imports, refactors without destroys, provider upgrades.
- Kubernetes in production Helm, RBAC, network policy, an ingress or mesh layer (Istio or Gateway API), and troubleshooting from kubectl down to the node.
- CI/CD you have designed, not just used: GitHub Actions (reusable workflows, OIDC / WIF, environments with required reviewers) and Cloud Build.
- SRE practice: SLOs and error budgets, on-call, incident command, postmortems, capacity and cost.
- Security as a habit: least privilege, secrets never in code or logs, audit trails — and the judgement to say "not without approval".
- Linux, networking and scripting: Bash and Python are daily tools.
- Written clarity: Our change process runs on pull requests and written decisions. You can state what you measured, what you changed and what you could not verify.
- Comfortable working remotely across time zones, with async communication as the default.
Nice To Have:
- HashiCorp Terraform Associate (or higher) certification.
- Google Cloud Professional Cloud DevOps Engineer or Professional Cloud Architect certification.
- Alibaba Cloud (ACK, OSS, KMS, VPC) — our DR platform.
- MongoDB Atlas, PostgreSQL, Redis or Kafka operations.
- Zero-trust network access and OIDC identity providers.
- Regulated-environment experience: data residency, financial-services controls, audit readiness.
- Experience running or reviewing LLM-based automation in production.
