Context:
Our EU-West BYOC environment relies heavily on the WebRTC softphone for agent interactions. Recently, agents have reported intermittent disconnections during peak hours, specifically when the call volume exceeds 500 concurrent sessions. These disconnections are not being accurately reflected in the Performance Dashboard. The agent status shows “Available” or “Talking” when the audio channel has actually dropped, leading to significant discrepancies in handle time and talk time metrics. This misalignment is causing issues with our WFM compliance reports and real-time monitoring capabilities. The issue appears to be isolated to the WebRTC implementation, as SIP trunk calls are logging correctly.
Question:
Is there a known correlation between WebRTC softphone stability and Performance Dashboard metric accuracy in the EU-West region? We need to understand if the dashboard is pulling status from the signaling layer or the media layer. If the media drops but the signaling remains active, how can we ensure the dashboard reflects the actual interaction state? We require a solution that allows for accurate reporting without manual intervention from the agents. Any insights into configuring the dashboard to detect these silent failures would be appreciated.