Your question is Monitor Service Health and Failures. 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).
You are responsible for a customer-facing service that handles payment-related traffic and depends on several internal and external systems. The service is healthy most of the time, but you have seen intermittent latency spikes, failed auth attempts, and occasional downstream timeouts. Your team wants a monitoring approach that helps you spot real incidents quickly without drowning in noise.
What tools or methods would you use to monitor system performance and detect meaningful failures? Explain how you would decide what to alert on, how you would reduce false positives, and how you would verify the monitoring is actually working.