Workforce Engagement Management

 View Only

Sign Up

  Thread closed by the administrator, not accepting new replies.

  • 1.  AI scoring includes ACD messaging in its evaluation of the question prompt.

    Posted 08-13-2025 06:13
    No replies, thread closed.

    Hi.

    Upon reviewing the AI scoring functionality, we are finding that ACD messaging is being included in the AI evaluation response. The transcription clearly understands it is not the agent speaking and that it is ACD, evidenced by the icons. However when the customer is thanked for holding before being connected to an agent, the AI response to an efficiency question seems to think the agent is repeatedly putting the customer on hold and therefore is not being efficient in their handling of the call. 

    The question is worded as 'Did the agent...', and as this forms part of the question prompt, I feel anything said by the ACD messaging should automatically be excluded, but it is not.

    Does this require us to disable 'Continue Voice Recording during Queue Wait' on the queue configuration to ensure it is not taken into consideration when using AI scoring?

    Any support or advice appreciated, Heather 


    #AIScoring(VirtualSupervisor)
    #QualityEvaluations
    #Recording,Policies

    ------------------------------
    Heather Henderson
    ------------------------------


  • 2.  RE: AI scoring includes ACD messaging in its evaluation of the question prompt.

    Posted 08-18-2025 10:07
    No replies, thread closed.

    Hi Heather! To avoid this there is an option in Genesys Cloud to suppress the call recording during the IVR. By doing so, there would be no transcript generated for the IVR portion of the interaction. Please see details here:   https://help.mypurecloud.com/articles/recording-in-genesys-cloud/



    ------------------------------
    Mark Fagus
    Senior Global WEM Business Specialist
    ------------------------------



  • 3.  RE: AI scoring includes ACD messaging in its evaluation of the question prompt.

    Posted 08-19-2025 03:00
    Edited by Heather Henderson 08-19-2025 03:04
    No replies, thread closed.

    Thanks Mark, we had already considered this, but it is very frustrating that the script recognises that it isn't the agent speaking, yet AI scoring does not. And the workaround suggested involves config change across all queues in order to get this to perform as we would expect. 



    ------------------------------
    Heather Henderson
    ------------------------------

    Original Message


Related Content