Predictive pacing clamps hard to 0.5x when the Quality evaluation sync throws a 502 BAD_GATEWAY on /api/v2/analytics/quality/evaluations. Running the 2024-12-18 release, US-East tenant. The outbound engine won’t let the multiplier climb past half-speed until that Quality API call returns 200 OK.
This happens every 45 minutes during peak outbound. WFM shows all agents as “Ready” in the adherence view, but the Quality module is apparently blocking the availability token refresh. The dialer assumes agents are unavailable for compliance reasons and just stops dialing.
Queue depth stays low, abandonment rate hits zero, but connect rate tanks. The campaign is set to predictive mode with a target answer rate of 30%. The logs show quality_sync_timeout in the campaign diagnostics. Restarting the campaign doesn’t fix it. Have to wait for the Quality service to recover. This is killing the Q1 outreach volume.
{
"error": "quality_sync_timeout",
"code": 502,
"pacing_clamp": 0.5
}
The save button on the Quality profile just spins when trying to toggle the integration off.