Your question is Observability, SLA/SLI/SLO. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
Explain SLA, SLO and SLI concepts and how you apply observability in your team's DevOps pipeline.
Asked in the System design (HLD) stage. First 20 minutes of the design round.
Discuss how you would define measurable reliability objectives for an ML-enabled service and its delivery pipeline. Explain which signals you would collect, how you would instrument each stage, how alerts and error budgets influence releases, and how you would detect model-specific issues such as feature drift and training-serving skew. Keep the answer practical and distinguish customer-facing commitments from internal engineering targets.