Does anyone know if there is a documented threshold for BYOC trunk latency that triggers silent failures in Architect flow routing? We are observing a discrepancy in our performance dashboards where conversations are marked as answered, yet the post-call survey logic within the Architect flow is not executing. The conversation detail view confirms the call connected successfully via the BYOC trunk, but the flow execution log shows the interaction terminating prematurely after the initial IVR node.
The environment is configured with strict adherence to European data residency requirements, and the BYOC trunk health metrics appear nominal. However, the queue activity reports indicate a spike in abandoned calls during peak European business hours (09:00-11:00 CET). This suggests that while the SIP signaling completes, the media path or control plane delay is causing the flow engine to timeout or drop the context before the survey node is reached.
We have verified the trunk configuration and the flow logic multiple times. The issue seems isolated to interactions routed through the specific BYOC provider rather than the standard Genesys Cloud telephony. Is there a known limitation regarding how the platform handles latency spikes on BYOC trunks that might interfere with flow execution? We need to determine if this is a configuration error on our end or a platform behavior we need to account for in our SLA reporting.