Use a measured example before changing the system.
1. Capture the slow operation: Record the action, start time, finish time, input size, and result. One concrete run is more useful than a general report that the system feels slow.
2. Compare it with the agreed baseline: Use the metric and environment defined in the handover, if the build included one.
3. Check volume and queues: A larger import, backlog, traffic spike, or storage limit can change processing time.
4. Check third-party services: Model providers, APIs, databases, and automation tools can add latency or throttle requests.
5. Check the infrastructure: Review the CPU, memory, storage, worker, and database signals available for that deployment.
Send support the measured example, deployment name, environment, expected baseline, and relevant log or request identifier. Investigation time and remediation are governed by the active support scope.