Skip to content
MVPeopleGroup
Back to insightsDevSecOps

What does a DevSecOps engineer do, and when to hire one?

Mathijs VisserNovember 15, 20265 min read
What does a DevSecOps engineer do, and when to hire one?

A DevSecOps engineer builds security into the software delivery pipeline. Every code change is scanned and checked automatically on its way to production, and the engineer's job is to make those checks accurate, quick and easy for developers to act on themselves.

So the role lives inside the CI/CD pipeline, where code is built, tested and deployed. The clearest way to describe it is to follow a single change from commit to production and look at what the engineer has set up at each stage.

What a DevSecOps engineer does, stage by stage

It starts on the developer's machine. Secret scanning catches a password or API key that has slipped into the code before it ever reaches the repository. Secrets are pulled from a vault at run time, so they never sit in configuration files. Branch protection and mandatory code review mean nobody can push to the main branch unseen.

At build time, two kinds of analysis run. Static application security testing (SAST) looks for vulnerable patterns in the code your team wrote. Software composition analysis (SCA) checks the open source libraries you pull in, code you did not write but are still shipping. The build should also produce a software bill of materials, or SBOM: an inventory of every component and version in the release. Generating and storing one with every build is routine DevSecOps work, and under the EU Cyber Resilience Act it will soon be a legal requirement.

Once the application is running in a test environment, dynamic testing (DAST) probes it from the outside, much as an attacker would. Container images are scanned for known vulnerabilities. This is where the hardest judgement call sits: which findings should break the build, and which should simply raise a ticket? Block everything and developers will find ways round the scanner. Block nothing and nobody reads the results.

Infrastructure is increasingly written as code, in Terraform, Bicep or Kubernetes manifests. Infrastructure-as-code scanning flags mistakes such as a publicly readable storage bucket or a container running as root before anything is provisioned. The engineer also hardens the pipeline itself. Who can trigger a release? What permissions do the runners have? Are deployment credentials short-lived?

After go-live the work carries on. A new vulnerability in a library you have used for months only surfaces if someone keeps matching the SBOM against fresh advisories and has agreed who picks up a critical alert, and how fast. I wrote about what happens when that ownership is missing in my piece on GitLab CVE-2026-97229, where the development platform itself became the target and an up-to-date version number turned out not to mean the patch was in place.

How it differs from application security

Job adverts often blur the two. An application security engineer focuses on the application: threat modelling at the design stage, reviewing sensitive code, triaging penetration test findings and teaching developers to write safer code. A DevSecOps engineer focuses on the pipeline around it and automates. Put simply, AppSec decides what needs checking and DevSecOps makes sure it happens on every change without anyone having to remember.

Smaller organisations often have one person covering both. If so, ask where they started. My rule of thumb is that a developer who moved into security tends to be stronger on code and design, while someone from operations or platform engineering usually builds the better pipeline. Our page on application security covers the first profile in more detail.

Which rules apply

Since 11 September 2026, manufacturers of products with digital elements, meaning software or hardware that runs software, have had to report under the Cyber Resilience Act (CRA). An actively exploited vulnerability in your product, or a severe incident, calls for an early warning to the national CSIRT within 24 hours, a full notification within 72 hours and, for a vulnerability, a final report no later than 14 days after a fix is available. In the Netherlands that CSIRT is the NCSC, the national cyber security centre, which states that company size makes no difference and that the duty also covers products already on the market before that date.

The CRA applies in full from 11 December 2027. From then on manufacturers must be able to identify and document the components in their products, including through an SBOM in a commonly used, machine-readable format. Knowing within 24 hours whether an exploited flaw affects your product is only realistic if that SBOM already exists and someone keeps it current. Keeping it current is a DevSecOps job.

Organisations in scope of the Dutch Cyberbeveiligingswet, the national law implementing NIS2 that has applied since 15 August 2026, have a second reason to care. Article 21 of the NIS2 Directive lists security in the acquisition, development and maintenance of systems, including vulnerability handling, alongside supply chain security. I would read that broadly enough to include your own build pipeline.

If you want a yardstick for your pipeline, NIST's Secure Software Development Framework (SP 800-218) is a sensible place to start. It sets out secure development practices that fit into any development lifecycle.

A gatekeeper is the wrong model

In my view, a DevSecOps engineer who spends most of the week holding releases back has got the job wrong. What you want is a developer seeing the problem a few minutes after committing, in their own tooling, with a suggested fix, and sorting it out while the code is still fresh in their mind.

A gatekeeper who reviews everything at the end turns into a queue. Teams work round the controls or switch scans off "until after the release". When you interview, ask how the candidate brings false positives down and how they win developers over. You will hear about their tooling along the way.

When an interim DevSecOps engineer makes sense

Interim suits work with a clear start and finish: building or overhauling a CI/CD pipeline, moving to the cloud or to a different development platform, or closing an audit finding on your development process by a fixed date. The CRA is another trigger. Setting up SBOMs, vulnerability management and a reporting process for your products is a project with a deadline.

Once the pipeline is in place, what remains is ongoing upkeep: tuning rules, onboarding new teams, following up alerts. For that you are better served by a permanent hire or a seconded specialist, with the interim engineer handing over properly before they leave.

In your request, spell out the platform (GitLab, GitHub or Azure DevOps), the cloud, the languages, the scanners already in place and what is going wrong. Ask for "a DevSecOps engineer with tooling experience" and you will get everyone's CV. Write "our SAST scan produces hundreds of findings and nobody looks at them" and the right people will recognise themselves. My checklist for a contractor request has more on this.

Looking for someone to sort out your pipeline or get your products ready for the CRA? Our page on hiring a DevSecOps engineer explains how we work, and through MVPeople we place interim security professionals. Or call or email me and we can work out what your pipeline needs.

Mathijs Visser

Mathijs Visser

Co-founder and Lead Consultant, MVPeople Group

Mathijs Visser founded MVPeople Group in 2022 and has recruited IT and cybersecurity professionals for Dutch enterprise organisations for more than 16 years. He specialises in interim CISO, IAM and GRC assignments.

Follow MVPeople Insights on Google

Add mvpeoplegroup.com as a preferred source to see our articles more often in Google's Top Stories and AI answers.

Want to know more?

Do you have questions about this article? Or would you like to discuss your specific situation? Get in touch with us, no strings attached.

READY TO FIND THE RIGHT CYBERSECURITY PROFESSIONAL?