Struggling to understand why the Agent Scripting API returns null for call_sid when correlating script events with our 15 APAC BYOC trunks. The standard voice analytics endpoint works fine, but the scripting metadata breaks down for 30% of sessions during peak hours. Our environment uses Genesys Cloud Engage with custom outbound routing logic. The issue seems tied to how SIP registration handles failover between primary and secondary carriers in the Singapore region. When a call switches trunks mid-session, the scripting engine loses the original SIP context. This creates gaps in compliance reporting for our QA team. We need accurate call_sid mapping to audit agent adherence against script steps. The problem persists across all 15 trunks, suggesting a systemic issue with how BYOC metadata propagates to the scripting service.
- Checked the
/api/v2/interactions/scripting/eventsendpoint and confirmedcall_sidis null for failed trunk handoffs. - Verified SIP credentials and failover logic in the trunk configuration, ensuring no registration timeouts are causing session drops.