Architect Flow Step Duration vs API Queue Metrics Discrepancy

Hey everyone, I’ve run into a really strange issue with the execution time metrics in our primary inbound queue flow within the EU-West environment. The specific issue involves a significant divergence between the ‘Average Execution Time’ reported in the Architect flow analytics and the wait times derived from queue analytics. The flow configuration is standard, yet the API data shows a 15-second lag that does not reflect in the dashboard views.

flow_id: f-8a9b7c6d-5e4f-3a2b-1c0d-9e8f7a6b5c4d
environment: EU-West
dashboard_view: Agent Performance
api_endpoint: queue analytics
metric: average_execution_time

What is the correct way to interpret this gap for reporting purposes?