DevOps Engineer

Webera builds and runs platforms for innovators with deadlines. Our DevOps On-Demand service keeps client infrastructure alive across AWS and Azure — CI/CD, Kubernetes, observability, incident response — for teams who need senior operations without hiring a senior operations team.

We are small on purpose. You will not be a ticket on a queue in someone else's process. You will own client environments end to end, talk to the client directly, and make calls without waiting for permission.

What you will actually do

• Run and improve Kubernetes workloads across AWS and Azure — day-2 operations, not just day-1 setup.
• Build and maintain CI/CD in GitHub Actions: pipelines that are boring, fast, and readable by whoever inherits them.
• Write and maintain infrastructure as code (Terraform), configuration (Ansible), and container builds (Docker) that hold up under review.
• Own observability — Prometheus, Grafana, logs, alerts that fire for real problems and stay quiet otherwise.
• Take incidents: triage, mitigate, communicate, then write the follow-up that prevents the repeat.
• Be a named face to the client. Explain what happened, in plain language, to people who are not engineers.

What we need from you

Technical

• Kubernetes in production, not in a tutorial. You can debug a pod that will not schedule, a service that will not resolve, and a node that is quietly full.
• AWS and Azure. Strong on one, competent and unafraid on the other.
• Docker: image hygiene, multi-stage builds, why the image is 1.2 GB and how to fix it.
• GitHub Actions: reusable workflows, secrets handling, self-hosted runners when it matters.
• Terraform or equivalent IaC. State is not a mystery to you.
• Linux and networking fundamentals that survive contact with a real outage.

The part that is harder to hire for

• Taste: You have opinions about what "good" looks like and you can defend them. You pick the simpler solution when it is enough, and you notice when a thing is ugly before a client does.
• Agency: You see a problem, you fix it, you tell us afterward. Nobody here is going to hand you a queue of tasks. Being blocked and quiet is the one failure mode we cannot work with.
• Human skills: You write clearly in English. You can sit in a call with a client's CTO, say "this was our mistake, here is what we changed," and be believed. You disagree with people without making it personal.

What we do not need

• Certifications as a substitute for scars.
• Someone who wants a fully specified backlog.
• Heroics. We are trying to make the on-call boring, not to celebrate the person who stayed up all night.

Location and hours

Remote, anywhere in Brazil. Core hours overlap with US Eastern business hours, because our clients are in the United States.

Contractor engagement (PJ). Real coverage is a rotation, not a person who is always awake. We build the schedule so nobody carries the night alone. Working weekends is not the plan; being reachable during your window is.

How the hiring works

1. Short written exchange — you tell us about a system you ran and what broke.
2. A working session, paid: a small, real problem from a scrubbed environment. No whiteboard algorithms.
3. A conversation with Joel about how you think and what you want to be doing in three years.
4. Reference call.

What we offer

• Direct work with US clients and the ownership that comes with a small team.
• No layers. You will talk to the founder every week.
• Room to shape the platform, not just consume it.
• Compensation: US$2,000–3,000 per month, paid in dollars, depending on what you bring. Tell us your number and we will tell you honestly whether we can meet it.

Interested? Send us the shortest possible note about something you fixed that nobody asked you to fix.