The business wants one global latency target across very different workflows. How would you respond?

Instruction: Describe how you would handle a single latency target that ignores workflow diversity.

Context: Tests how the candidate diagnoses the problem, chooses the safest next step, and reasons through recovery. Describe how you would handle a single latency target that ignores workflow diversity.

Official answer available

Preview the opening of the answer, then unlock the full walkthrough.

I would push for a small set of workflow-specific targets instead of one global number. Different flows have different tolerance for delay and different structural causes of latency, so one blended target usually creates the wrong incentives.

For example, a quick Q and A interaction may need a...

Related Questions