Troubleshoot a production incident end-to-end

You are on call for a microservices-based banking platform and users report intermittent failures and elevated latency on transfers starting at a specific time. Walk through your troubleshooting approach end-to-end: which dashboards, metrics, logs, and traces you would inspect first; how you would form and test hypotheses; how you would isolate whether the issue is in the client, network, service, database, or external dependency; what short-term mitigations you would apply safely; and how you would verify the fix and drive postmortem actions to prevent recurrence.