DevOps EngineerDevOps Engineer

Tech stack

Kubernetes and Terraform are the core — that pair is on nearly every engagement, and it's what the assignment round is built around. The rest moves with the client.

• Kubernetes — managed (EKS/GKE/AKS) and, more often than you'd expect, on-premise, multi-cloud and air-gapped.
• Terraform — every environment reproducible from an empty account, plus Helm and GitOps with Argo CD or Flux.
• AWS, GCP and Azure. Deep in one is the baseline; we work across all three and you'd be expected to get productive in the others.
• CI/CD tooling — GitHub Actions, GitLab CI, or equivalent — built around signed images, SBOMs and scanned dependencies.
• OpenTelemetry, Prometheus and Grafana for the parts that tell you what's actually happening.
• Go or Python for the internal tooling and automation that glues it all together.

What you'd build

• The platform a client's product runs on: clusters, deploy paths, environments, the glue that makes a handover survivable.
• CI/CD pipelines developers actually trust, with supply-chain hygiene built in from the start.
• Observability that answers questions rather than filling dashboards — OpenTelemetry, Prometheus, structured logs, meaningful SLOs.
• Infrastructure-as-code for every environment, so nothing depends on a console click only one person remembers making.
• Controls that hold up to SOC 2 and ISO 27001 audits without a scramble.
• On-call runbooks and incident response that leave the client's team more capable than they found it, not more dependent on us.

What we're looking for

3-5 years in a DevOps, SRE or infrastructure engineering role, with production systems you owned rather than just touched. A degree doesn't matter — we don't ask. Proof beats a CV: a repo, a cluster you've run, an incident you've handled that people depended on.

We do have a preference: people who have worked at startups. Early-stage work teaches a particular thing — owning a problem end to end with nobody to hand it to — and that is most of this job.

• 3-5 years of hands-on infrastructure or DevOps experience, most of it in production, not staging.
• Work at real Kubernetes depth — not just kubectl apply. Scheduling, networking, RBAC, CRDs, and what happens when a node goes away.
• Reach for Terraform (or equivalent IaC) by default, not as a last resort.
• Move fluently across AWS, GCP and Azure — deep in one, dangerous in the other two — with pipelines, releases, IAM, networking and cost all owned by you.
• Build CI/CD that a team actually trusts, with security baked into the pipeline rather than bolted on after.
• Bring enough Linux and networking to debug below the abstraction, not just around it.
• Run something in production and stay responsible for it — you've carried a pager for something that mattered.
• Apply security instincts by default: least privilege, secrets handling, tenancy boundaries.
• Write clearly. Runbooks and design docs are half of what we deliver.
• Have worked with remote startups across time zones, and know what that actually takes: writing things down, deciding without a meeting, and being reachable when it counts without being online at 3am.
• Learn fast. Clients arrive with stacks we've never touched, and the useful skill is getting to a confident change in that infrastructure within days.
• Hold your own in front of a client: scoping, disagreeing well, and saying “we can't measure that yet” out loud.

Nice to have

• Go or Python for building internal tooling, operators or automation.
• Multi-tenant SaaS platform work, or a compliance audit you've been through.
• Open-source contributions in the cloud-native ecosystem.
• GPU scheduling or inference-serving infrastructure experience.
• Early-stage startup experience — founding team, first engineers, or your own thing that shipped.

How to apply

Email [email protected] Email us with:

• Your CV or LinkedIn.
• A short note — a few paragraphs is plenty — on infrastructure you built and ran, what broke, and what you changed because of it.
• Links to anything public: repos, writing, talks, side projects, homelab clusters — whatever shows the work.
• Your location and time zone, and how far out you'd be available.
• A line on why you, for this role specifically. That's the part we read closest.

We'll keep your details on file for the projects this posting is aimed at and come back to you when one lands. If you'd rather we didn't hold them, say so and we'll delete them.