ML engineer behavioural interview questions

As asked

Walk me through the biggest production incident you've personally been on the front line of. What happened, what did you do, and what changed afterwards?

Sample answer outline

Use the STAR structure but lean into the technical specifics. Briefly describe the impact (users affected, revenue, duration). Walk through detection (what alerted you and how), diagnosis (what you ruled out, the dead-end paths), mitigation (the actual fix), and the postmortem outcome (what changed in the system or process). Interviewers look for ownership, calm under pressure, root-cause thinking (not 'we restarted it'), and the discipline to convert pain into permanent fixes.

Expect these follow-ups

• What would you do differently if you saw the same symptoms tomorrow?
• Whose fault was it? How did you handle that conversation?
• Did the postmortem actions actually land, or did they slip?

As asked

You have a quarter to plan. Engineering wants to pay down infrastructure debt; product wants to ship three big features. How do you make the call?

Sample answer outline

Reject the framing as binary - good engineering leaders do both. Quantify the debt: where is it slowing new feature work, where is it a reliability or security risk, where is it just aesthetic. Talk to product about which features are time-sensitive (a launch tied to a market window) vs which can slip a quarter. Propose a split, e.g. 70/30 with the debt items chosen specifically to unblock the features. Show your work to both sides so neither feels ignored.

Expect these follow-ups

• What if the CTO says 'no debt work this quarter, all features'?
• How do you measure whether the debt work actually paid off?
• Give an example where you were wrong about which side to prioritise.

As asked

Tell me about a time something broke in a system you were working on and you helped get it fixed. It does not need to be a major outage. What happened and what did you take away from it?

Sample answer outline

An earlier-career version of the incident story. Pick a real, smaller event: a deploy that broke a page, a job that silently stopped, a config that was wrong. Use a light STAR structure: what the symptom was, how you noticed or were told, what you checked, and the fix. Interviewers at this level want to see curiosity, that you did not just guess and restart, and that you learned something concrete, even if the learning is as simple as adding a check or asking for help sooner. Honesty about what you did not know is fine here.

Expect these follow-ups

• How did you confirm your fix actually solved the problem?
• What would you check first if you saw the same symptom again?
• Who did you ask for help, and when?

As asked

Tell me about an incident where the real cause was not one bug but a systemic weakness: an architectural assumption, an organisational gap, or a slow-burn failure that several safeguards missed. How did you drive the fix beyond the immediate mitigation?

Sample answer outline

This is the senior version: the interviewer wants systems thinking and influence, not just firefighting. Set the impact concretely, then show that the immediate mitigation was the easy part. Walk through the contributing factors, plural: why detection was late, why a safeguard did not trigger, what assumption in the design no longer held at the current scale. The strong answer demonstrates that you drove durable change across team boundaries: a new invariant enforced in code, a class of failure designed out, an ownership gap closed. Be candid about what was hard organisationally and which follow-ups slipped, and how you held the line on the ones that mattered.

Expect these follow-ups

• Which contributing factor was hardest to get the organisation to accept?
• How did you prevent the action items from quietly slipping?
• What design assumption broke, and how did you make the new one explicit?

As asked

Tell me about the biggest incident you handled that lived in the platform layer: a broken deploy pipeline, an infrastructure change gone wrong, a certificate expiry, or a cluster failure. Walk me through detection, mitigation, and what changed in the platform afterwards.

Sample answer outline

Frame it through a platform lens, where the blast radius is every team that depends on you. Describe the impact across consumers, not just one service. Detection: what alerted you, and whether it was your monitoring or a downstream team that noticed first. Mitigation: the rollback or break-glass procedure, and whether it existed before the incident or had to be improvised. The strong answer ends with platform-level prevention: a guardrail in the pipeline, a pre-deploy check, an expiry alert, automated rollback. Interviewers listen for ownership of shared infrastructure and the discipline to turn one painful event into a control that protects every team.

Expect these follow-ups

• Did a self-service guardrail exist, or did you have to build one after?
• How did you communicate with the many teams affected at once?
• What pipeline or infrastructure check would have caught this earlier?