WebRTC softphone audio dropouts causing false adherence flags in Chicago region

Voice & Telephony
1

Why does this setting in the WebRTC softphone configuration cause intermittent audio dropouts that are being flagged as “off-phone” events in our schedule adherence reports?

We are seeing a consistent pattern where agents using the Genesys Cloud WebRTC softphone (v1.2.4) in the America/Chicago timezone experience 1-2 second audio gaps during high-volume periods. These gaps are not perceived by the customer as disconnects, but the system is logging them as off-phone intervals, which negatively impacts our adherence scores. This is particularly problematic for our weekly schedule publishing process, as we rely on accurate adherence data to adjust future shifts and agent preferences.

Here are the steps to reproduce the issue:

  1. Log in as an agent with a standard WFM schedule in the America/Chicago region.
  2. Initiate an inbound call via the WebRTC softphone.
  3. Engage in conversation for at least 5 minutes.
  4. Observe the adherence report in real-time or via the WFM API endpoint /api/v2/wfm/schedules/agent-adherence.
  5. Note the sudden “off-phone” flag appearing despite the agent actively speaking.

We have already checked the network conditions for these agents, and the jitter and packet loss metrics are well within acceptable limits (<2% packet loss, <30ms jitter). The issue seems to be isolated to the WebRTC client’s handling of audio streams during peak traffic times. We are also using the latest version of the Genesys Cloud API for our WFM integrations, so we suspect this might be a client-side issue rather than an API limitation.

Has anyone else encountered similar audio dropout issues with the WebRTC softphone that affect adherence metrics? We are looking for a workaround or a configuration change to prevent these false flags from impacting our schedule adherence reports. Any insights or troubleshooting steps would be greatly appreciated.

2

Make sure you:

  • Enable enableWebRtcAudioLevelDetection in the softphone config to ignore low-volume gaps
  • Set audioActivityTimeout to 3000ms to prevent premature off-phone flags
3

You need to adjust the WebRTC configuration parameters to prevent false adherence flags caused by transient audio silence. The previous suggestion regarding audioActivityTimeout is a good start, but the root cause often lies in how the softphone interprets jitter buffers during peak load.

  • Set enableWebRtcAudioLevelDetection to true to ensure the system distinguishes between silence and actual disconnection.
  • Increase audioActivityTimeout to 3000ms or higher. This buffers against the 1-2 second gaps common in high-volume periods without triggering immediate “off-phone” status.
  • Verify that webrtcAudioLevelThreshold is not set too high. A lower threshold helps detect subtle audio activity that might otherwise be missed during network congestion.

These adjustments should stabilize the adherence reporting. If issues persist, check the network latency metrics for the Chicago region, as excessive jitter can exacerbate these false negatives. Monitoring the SIP signaling logs for any 408 timeouts during these dropout events can also provide further insight into whether the issue is purely client-side or involves trunk-level latency.