DevOps and SRE resume keywords, counted from real job postings

We read postings directly from employers' applicant tracking systems (Greenhouse, Lever, Ashby, Workday and others). For each role we take the newest live postings whose title matches the role, remove re-posts of the same ad, and scan only the description. Benefits, salary and equal-opportunity paragraphs are ignored, and general words like "communication" only count when they appear in the responsibilities or requirements, not in the company's description of itself.

Each skill is matched with its common synonyms (k8s for Kubernetes, "Amazon Web Services" for AWS, GitLab CI or Jenkins as separate tools) and counted once per posting, however many times the ad repeats it. The percentage is the share of postings that mention the skill at all. The role's own name ("DevOps", "Site Reliability Engineering") is left out, because every posting says it.

This is the same keyword matcher our free CV score checker uses, so the terms below are the ones it will look for in your CV.

Reading the table: anything above roughly half of postings is close to a baseline expectation for the role. In the current data that includes CI/CD, Kubernetes, Terraform, Infrastructure as Code, Python, AWS. Kubernetes appears in 68% and Terraform in 61% of postings, so if you have either, make sure it's visible in both your Skills section and a bullet.

The full list, with a copy button, is on the DevOps engineer resume keywords page.

SRE postings put more weight on running systems than on building pipelines: incident response, observability, SLOs and on-call show up alongside the infrastructure tools. If you've carried a pager, say so, with the scale (services, traffic, team size).

The full list is on the site reliability engineer resume keywords page.

The two titles overlap a lot, and many companies use them interchangeably. The differences are in emphasis, and they're worth reflecting in your summary line when you apply for one or the other:

Skills SRE postings mention noticeably more often than DevOps postings:

• DevOps: 48% of SRE postings vs 0% of DevOps postings
• Incident management: 75% of SRE postings vs 27% of DevOps postings
• SLOs: 43% of SRE postings vs 0% of DevOps postings
• Distributed systems: 32% of SRE postings vs 0% of DevOps postings

And the reverse, skills DevOps postings lean on more:

• Git: 27% of DevOps postings vs 0% of SRE postings
• Agile: 24% of DevOps postings vs 0% of SRE postings
• CI/CD: 83% of DevOps postings vs 60% of SRE postings
• Site Reliability Engineering: 20% of DevOps postings vs 0% of SRE postings

A practical rule: for DevOps roles, lead with what you automated (pipelines, provisioning, developer tooling). For SRE roles, lead with what you kept reliable (availability targets, incident counts, latency, cost).

Tools aren't the whole list. In the requirements sections of the same DevOps postings, communication (39%), bachelor's degree (27%) and collaboration (26%) come up most. These rarely belong in a Skills list; they belong in bullets.

"Collaboration" in a DevOps ad usually means working with application teams: running a migration with them, building a self-service tool they adopted, or pairing on their first production deploy. "Communication" usually means writing: runbooks, postmortems, RFCs and status updates. One bullet that shows each beats any adjective.

Seniority shows up the same way. Staff and senior postings mention mentoring and leading initiatives far more than tool depth, so if you've onboarded engineers, led an on-call rotation or driven a platform migration across teams, give it its own bullet with the scope.

1. Only list what you've used. Interviews for these roles are hands-on. A tool you can't discuss for five minutes costs you more than it gains.
2. Use the exact name. "Terraform", not "IaC tools"; "GitHub Actions", not "CI". Recruiters search for the product name.
3. Show each core skill in a bullet, not just the Skills list: what you built or ran with it, and one number.
4. Keep your Skills section grouped the way it already is (Cloud, IaC, CI/CD, Observability, Languages). Add missing terms to the right group instead of a new "keywords" block.
5. Match the cloud. If the ad says AWS and you've used AWS, name the services (EKS, Lambda, IAM). If you've only used GCP, say so; cloud skills transfer, but don't imply otherwise.

These are illustrations of the wording, not real candidates. Every "after" only uses facts already in the "before".

Notice what changed: the tool names are explicit, there's a scope or number, and the verb says what you did. Nothing new was claimed. If your "before" doesn't contain the facts for an "after", leave it as it is.

Certifications appear in a minority of DevOps and SRE postings and are almost never the deciding factor for experienced candidates. They help most when you're changing careers or moving to a new cloud: a CKA or an AWS certification is evidence you can't easily show otherwise. List them in their own short section with the year. Check the role pages for the live share of postings that mention each one.

Before you apply, check your CV against the actual job: our free CV score checker shows which of the ad's keywords your CV already covers and which are missing. To make sure a remote role can hire you where you live, use the remote eligibility checker.