Ran into a weird issue today with the WFM Integration block in Architect. The flow retrieves schedule status correctly for 90% of agents, but returns a 400 Bad Request for specific users during shift swaps. The payload looks identical to successful requests, just different agent IDs.
The error response body contains "error_summary": "Invalid schedule reference". This happens immediately after publishing the weekly schedule via the API. Anyone see this when WFM data propagates to the IVR?
The best way to fix this is to introduce a 30-second delay. WFM data propagation is not instantaneous after publishing. The 400 error indicates the schedule reference does not yet exist in the query index.
The easiest way to fix this is adding that delay. In my JMeter tests, querying immediately after publish fails because the index hasn't updated. The 400 error confirms the reference is missing. Wait for propagation to complete before the IVR hits the WFM block.