Genesys Cloud - Main

 View Only

Sign Up

Experience with auto-answer in Genesys CX Cloud for Salesforce: timing between Omni events, acceptCall, and WebRTC

  • 1.  Experience with auto-answer in Genesys CX Cloud for Salesforce: timing between Omni events, acceptCall, and WebRTC

    Posted 10 days ago
    Hello everyone,
    We are reviewing a voice interaction flow using Genesys Cloud CX integrated with Salesforce Service Cloud Voice through CX Cloud for Salesforce. The implementation uses auto-answer on the Salesforce/SCV side, and we are trying to better understand the latency between Salesforce Omni-Channel events and the Genesys/WebRTC connector events.
    From the browser console logs, the overall sequence generally looks like this:

    Genesys conversation alerting
    Salesforce SCV createVoiceCall
    Salesforce Omni OMNI.WORK.ASSIGNED
    Salesforce Omni OMNI.WORK.ACCEPTED
    Connector acceptCall
    acceptPendingWebRTCSession
    WebRTC session started
    Media started
    ICE connected
    Call connected

    I would like to hear from anyone who has implemented or troubleshot this flow in production:

    What has been your experience with the timing between createVoiceCall and OMNI.WORK.ASSIGNED?
    When auto-answer or auto-accept is enabled, does OMNI.WORK.ACCEPTED usually happen almost immediately after OMNI.WORK.ASSIGNED?
    After OMNI.WORK.ACCEPTED, does acceptCall usually trigger within a few milliseconds?
    In production environments, where have you seen the biggest latency contribution: Salesforce Omni-Channel, VoiceCall creation, the Genesys CX Cloud connector, or WebRTC/browser media setup?
    Which configuration changes helped the most? Presence Configuration, Omni routing, SCV settings, console performance, browser/device settings, permissions, loaded console components, or something on the Genesys side?

    It would also be helpful to understand if anyone measures these steps using a timeline like this:

    T0 = Genesys alerting
    T1 = createVoiceCall start
    T2 = createVoiceCall end
    T3 = OMNI.WORK.ASSIGNED
    T4 = OMNI.WORK.ACCEPTED
    T5 = acceptCall start
    T6 = acceptCall end
    T7 = WebRTC session started
    T8 = Media started
    T9 = ICE connected / call connected

    What is the lowest end-to-end time you have been able to achieve, and which part of the flow was the hardest to optimize?
    Thanks in advance for sharing any real-world experience, metrics, or recommendations.

    #API/Integrations
    #ArchitectandDesign
    #Implementation
    #Other

    ------------------------------
    Att,
    Breno Canyggia Ferreira Marreco
    ------------------------------