DevOps Engineer

We're hiring a DevOps Engineer with 4-5+ years of experience running production cloud infrastructure who has built at least one environment from scratch. You will own how our environments are provisioned, deployed, and kept healthy, using Kubernetes, Helm, Terraform, and ArgoCD.

Responsibilities:

• You will make every environment reproducible from code, from the first VPC to the running application.
• Stand up new environments end-to-end (network, Kubernetes cluster, databases, DNS, TLS, secrets, and deploy pipeline) with Terraform, not by hand.
• Bring existing hand-built environments under Terraform and Helm so any of them can be rebuilt from the repository
• Run our Kubernetes clusters: node pools, autoscaling, resource requests and limits, ingress, and version upgrades.
• Write and maintain Helm charts for application services, background workers, and their stateful dependencies, with clean per-environment values.
• Own CI/CD: build container images, promote releases from dev to staging to production, and roll back safely.
• Manage short-lived preview environments for pull requests, including cleanup when a PR closes.
• Set up monitoring, logging, and alerting; respond to infrastructure incidents; and write the follow-ups.
• Manage access and secrets with least-privilege IAM, rotated credentials, and no secrets in code.
• Keep cloud spend visible and right-sized.
• Write runbooks so any engineer can deploy, debug, or recover an environment.

Requirements:

• Experience: 4-5+ years managing cloud infrastructure (AWS, GCP, or Azure).
• Core stack: Kubernetes, Helm, Terraform, Argo CD.
• Must have done: taken at least one environment from an empty cloud account to production.
• All of these are required; the first two are the bar we screen on.
• 4-5+ years managing production infrastructure on at least one major cloud: AWS, GCP, or Azure.
• At least one environment built from scratch and taken to production, covering networking, compute, a Kubernetes cluster, databases, DNS/TLS, and CI/CD. You can explain the decisions you made and what you would change.
• Kubernetes in production: deployments, services, ingress, autoscaling, resource tuning, cluster upgrades, and debugging pods that are pending or crash-looping.
• Helm: You have written charts and not only installed them, including templates, per-environment values, upgrades, and rollbacks.
• Argo CD (GitOps): You have run Argo CD in production, with every environment synced from Git and releases promoted and rolled back through commits, not manual kubectl changes.
• Terraform: You have written and maintained modules, used remote state with locking, run plans and apply them in CI, and imported existing resources. CI/CD with GitHub Actions, GitLab CI, Jenkins, or similar, building and deploying container images.
• Docker and Linux fundamentals, plus scripting in Bash or Python.
• Networking: VPCs, subnets, load balancers, DNS, TLS certificates, and private connectivity between services.

Nice to have:

• None of these is required, but each one shortens your ramp-up.
• ML Ops or AIOps: deploying and serving ML or LLM workloads (model serving, GPU node pools, model versioning), or using AI-driven tooling for alerting and incident response.
• GCP specifically: GKE, Cloud SQL, Artifactory Registry, and IAM.
• Running stateful services on or alongside Kubernetes, such as PostgreSQL, Redis, Neo4j, and queue-backed workers.
• Observability with Prometheus, Grafana, Loki, or ELK, and OpenTelemetry.
• Secrets management with Vault, External Secrets, or a cloud secret manager.
• Running separate, single-tenant environments for individual customers.
• Certifications such as CKA, CKAD, HashiCorp Terraform Associate, or a cloud provider's associate or professional track.