A DevOps engineer builds and runs the systems that get your code from a developer's laptop into production and keep it running there: your cloud infrastructure, your build and deployment pipelines, and your monitoring. In a small team those jobs start out shared among the developers. A DevOps engineer is worth bringing in when they stop being side tasks and start slowing every release down. This guide covers what the role does day to day, how it differs from the titles it gets confused with, and the signs your team has reached that point.
What a DevOps engineer does day to day
The work falls into three parts.
- Infrastructure as code. Servers, databases, networks and permissions on AWS, Google Cloud or Azure, defined in code (Terraform, Pulumi, CloudFormation and the like) rather than clicked together in a console. Infrastructure written as code can be reviewed, repeated and rebuilt, and you can see who changed what.
- Build and deployment pipelines. The CI/CD pipelines that test every change and deploy it the same way every time, so a release is a routine event rather than a careful ceremony.
- Monitoring and alerting. Metrics, logs and alerts on the things that matter to your customers, so you hear about a problem from a dashboard rather than from a support ticket.
Around those three sit the jobs nobody else owns: cost (a cloud bill that grows faster than your traffic), access (who can reach production, and how that is audited), and security basics such as patching and secrets management.
DevOps engineer vs cloud engineer vs SRE
The titles overlap, and many job ads use them interchangeably.
| Title | Usually emphasizes |
|---|---|
| DevOps engineer | Pipelines, automation and the path from code to production |
| Cloud engineer | Designing and running infrastructure on a cloud provider |
| Site reliability engineer (SRE) | Reliability targets, incident response and capacity, often at larger scale |
| Platform engineer | Internal tools that let developers deploy and run services themselves |
For a small team the distinctions rarely matter. What matters is the work you need done, so describe the work, not the title.
Signs your team needs a DevOps engineer
- Deploys are manual, or only one person knows how to do them. If a release depends on one developer's memory, it depends on their availability too.
- Nobody can rebuild your infrastructure. If the production environment was assembled by hand in a cloud console, it cannot be recreated reliably after a mistake.
- Your developers spend real time on infrastructure. Every hour a product developer spends wrestling with a pipeline is an hour not spent on the product.
- You find out about outages from customers.
- Your cloud bill keeps growing and nobody knows why.
- A customer or auditor has started asking how production access is controlled.
If none of these is true yet, your developers can probably carry the work for now. If several are, a DevOps engineer is likely worth bringing in.
Where to start
A useful first deliverable is narrow and concrete: one service's infrastructure written as code and deployed through a pipeline, with alerts on the metrics that matter to you. It proves the approach on something real, and the rest of your systems can follow the same pattern.
Mid-level or senior
Choose senior if nobody on your team will set direction or review the work: a senior engineer has owned this kind of work alone before, which is the usual case for a team's first DevOps engineer. Choose mid-level if a senior engineer on your team will set direction and review the changes.
Giving a contractor access to production
Infrastructure work means access to production, so decide the boundaries before day one: which accounts and environments the work needs, how changes are reviewed before they apply, and how access is removed when the engagement ends. Under our agreements, if a contractor materially breaches your security or confidentiality requirements, acts dishonestly or acts unlawfully, you can have them removed immediately: their access is suspended and their fee stops that day (MSA §6). The infrastructure code they write is yours as it is created (MSA §3).
Frequently asked questions
Is DevOps a role or a practice?
Both. DevOps began as a practice, developers and operations sharing responsibility for running software, and the title grew up around the people who build the tooling that makes that practice work.
Does a startup need a DevOps engineer?
Not at the very start, when a managed hosting platform and a simple pipeline cover it. Once deploys, infrastructure and monitoring take real time away from product work, it usually does.
What is the difference between DevOps and a system administrator?
A system administrator traditionally keeps existing servers running. A DevOps engineer automates how infrastructure is built and how code reaches it, mostly in code rather than by hand.
Which cloud should a DevOps engineer know?
Yours. Name the cloud and the tools you run when you describe the role, and candidates are measured against that.
Where to go from here
sourceBOLD engages mid-level and senior DevOps and cloud engineers across the Americas, month to month, as their contractor of record. DevOps and cloud shows typical starting prices, and how we vet candidates covers what happens before you meet anyone.