Interview scorecard
DevOps Engineer
Score each skill from 1 to 5. A 1 is below the bar for this role, a 3 meets it, and a 5 is above it. The examples show what each sounds like. Use 2 and 4 for answers that fall in between.
Infrastructure as code
Builds and changes infrastructure through versioned, reviewed code instead of manual console changes.
Questions
- Tell me about infrastructure you defined in code. What tools did you use and why?
- Tell me about a time a code change to infrastructure went wrong. How did you recover?
- How do you keep several people's infrastructure changes from clashing or drifting from what's running?
1: Below the bar
Describes mostly manual console changes. Can't explain how code changes are tracked, reviewed or reversed.
3: Meets the bar
Describes reusable, version-controlled code with reviews. Handles drift and rollback reasonably but leans on team habits more than checks.
5: Above the bar
Describes tested, reviewed, reusable modules with automated drift checks. Explains how they roll back safely and share state across a team.
CI/CD pipelines
Builds and maintains automated pipelines that test and ship code safely and quickly.
Questions
- Tell me about a pipeline you built or improved. What problem were you solving?
- Tell me about a time a deployment failed. How did your pipeline help or hinder?
- Tell me about a time you made a slow or flaky pipeline faster.
1: Below the bar
Only describes running existing pipelines. Cannot explain what each stage does or why it is there.
3: Meets the bar
Describes building or changing a pipeline with clear stages. Handles failures and slowness but mostly after they cause problems.
5: Above the bar
Explains stage choices and trade-offs, with rollback and speed built in early. Backs claims with numbers like build time or failure rate.
Monitoring and alerting
Chooses what to measure and sets alerts that tell the team about real problems quickly.
Questions
- Tell me about a service you set up monitoring for. What did you choose to track?
- Tell me about an alert that was noisy or missed a real problem. What did you change?
- How do you decide whether an alert should wake someone up at night?
1: Below the bar
Tracks only basic server numbers like CPU. Alerts fire on everything or nothing, and they can't say how they'd fix that.
3: Meets the bar
Tracks sensible measures for the service and has tuned alerts after noise or misses. Links alerts to impact only some of the time.
5: Above the bar
Ties measures to user impact, sets clear rules for what pages someone, and cut noisy alerts with evidence. Reviews alerts regularly.
Incident response
Restores service calmly during outages, communicates clearly, and follows up so the same failure does not repeat.
Questions
- Tell me about a production outage you handled. What did you do first?
- Tell me about a time you updated others during an incident. What did you say?
- Tell me about an incident review you took part in. What changed afterward?
1: Below the bar
Describes guessing at fixes or waiting for others. Shares no updates and can't name any change made afterward.
3: Meets the bar
Follows a clear sequence to restore service and keeps people informed. Takes part in a review and names a follow-up.
5: Above the bar
Restores service first, then finds the cause. Gives regular plain updates and drives fixes that prevent repeats.
Security awareness
Spots security risks in everyday infrastructure work and builds protections in before problems appear.
Questions
- Tell me about a security risk you found in infrastructure you worked on. What did you do?
- How do you keep passwords, keys and access tokens safe in your pipelines and servers?
- Tell me about a time security and delivery speed pulled against each other. What did you decide?
1: Below the bar
Treats security as someone else's job. Stores secrets casually and can't describe a risk they ever spotted or fixed.
3: Meets the bar
Describes sensible habits like secret stores and least access. Fixes risks when they're pointed out, but rarely hunts for them.
5: Above the bar
Looks for risks unprompted, builds checks into daily work, and explains how they weighed risk against speed with a clear reason.
Overall recommendation
Decide on your own, before you compare notes with the rest of the panel.