Engineering and IT · Job description template
DevOps Engineer job description template
As a product grows, someone has to own how code reaches production and what happens when production breaks. The DevOps engineer runs the deployment pipeline, the servers and the monitoring. They are accountable for releases being routine and outages being rare and short.
First-year outcomes
What should be visibly different a year from now. Results, not tasks.
- Every service deploys to production through an automated pipeline with one approval step, and no one deploys by logging in to a server.
- Database backups are restored to a test environment every month, and a full restore takes under two hours.
- Uptime for the customer-facing product is at least 99.9 percent over the year, measured by an external uptime monitor.
- All servers, networks and DNS records are defined in code, such as Terraform, so any environment can be rebuilt from the repository.
The work
Recurring work
- Maintain the build and deployment pipelines and help developers when a pipeline fails.
- Watch dashboards and alerts, and lead the response when production has a problem.
- Apply security patches and renew certificates before they expire.
- Review cloud costs each month and remove unused resources.
Projects
- Move any manual server setup into infrastructure as code.
- Set up a monthly backup restore test and write the runbook.
- Tidy up alerts so each one points to a real problem and a clear first step.
Decisions they own
- How environments such as staging and production are set up and separated.
- Which alerts wake someone up and which wait until morning.
- When to roll back a deployment during an incident.
People they work with
- Developers, to make the pipeline fast and to agree how services are deployed.
- The engineering lead or CTO, about incidents, risks and cloud spend.
- Support staff, to share status during an outage.
Requirements
Essential on day one
- Has run production services on a cloud provider such as AWS, Azure or Google Cloud, or on managed servers.
- Linux administration and shell scripting.
- Has built or maintained a CI/CD pipeline, for example with GitHub Actions or GitLab CI.
- Working knowledge of containers such as Docker, networking basics and DNS.
Helpful, but can be learned
- Your specific cloud provider, if they know another well.
- Kubernetes, if your setup uses it.
- Your monitoring and logging tools.
How to assess it
One piece of evidence for each first-year outcome.
| Outcome | Evidence to look for |
|---|---|
| Automated deploys with no server logins | Ask them to draw a pipeline for a web service from commit to production. Good answers include tests, a build artifact, staging, an approval step and a rollback path. |
| Monthly tested restores under two hours | Ask about the last time they restored a backup for real. Good candidates describe the steps, the time it took and what they fixed in the process. |
| 99.9 percent uptime | Walk through a mock outage where the site returns errors. Good candidates check recent changes first, communicate status and roll back before deep debugging. |
| Infrastructure defined in code | Show a short Terraform or similar file and ask what it creates and what could go wrong. Good answers spot hard-coded secrets or missing state locking. |
Opening for the job ad
You will own how our code reaches production and how we keep it running. In your first year, deploys should be automatic and boring, backups should be proven by real restores, and the whole setup should be rebuildable from code.
Turning this into an ad? See how to write a job ad that gets applicants.
Next: interview for these outcomes
Eight structured questions for this role, each tied to an outcome above.
DevOps Engineer interview questionsSimilar roles
Your next hire, start to finish, in one place.
Free workspace. No card. Full scoring free for 7 days.