Adding a suggested case

Consider a scenario where a customer calls the support center and states that he has an unrecognized charge on his credit card. The expected outcome is that Pega Voice AI suggests the Dispute transaction case to the CSR.

To accomplish this outcome, log in to Pega Customer Service as an administrator, and then add the case to the Voice AI channel.

You open the Voice AI channel.

On the Configuration tab, in the Suggested cases section, click Add case type.

In the Response configuration dialog box, you select the Dispute transaction Case Type, then add a Create case command.

The create case command is a topic added to the channel, for example, Dispute transaction. This topic is associated with the Dispute transaction Case Type.

On the Text analysis tab, you configure words or short phrases that a customer might use to identify the topic. For example, you can add the keywords wrong, invalid, transaction as words that the Voice AI channel uses to trigger the Dispute transaction case.

These keywords are a quick way to "feed" words to the model. As Pega Customer Service runs cases, the NLP model learns and associates more words with the Case Type. You can access the associated models in the Prediction Studio, which provides more detailed configuration settings; typically, a a data scientist in your organization monitors these models.

You save your channel configuration, then test the new case.

Use the Preview console to test the keywords. Enter wrong and invalid in the console, and then click Send.

The Preview console confirms that Pega Voice AI detects the Dispute transaction topic.

Consider the following scenario to learn how the case works with a live call.

Bob is a CSR who handles inquiries about credit and finance issues. Bob logs in to the Interaction Portal. He has an incoming call from a customer, Sara Connor. Bob accepts the call and asks how he can help. Sara says, "I have an unrecognized charge on my credit card." Bob sees that the Interaction Portal suggests a Dispute transaction case. He opens the case, and then continues the call with Sara.

Consider a scenario where a customer calls the support center and states that she wants to update her address. The expected outcome is that Pega Voice AI suggests the Address change case to the CSR.

You add the Address change case type. On the Text analysis tab, you describe the topic by entering a sample customer request. When a topic is not easily identified by specific keywords, you can describe the topic and provide some example utterances that a customer might use that would lead the CSR to open a specific Case Type. Pega GenAI can generate data based on your topic description.

You select Generate utterances. Pega GenAI creates more utterances based on the topic description you entered.

You can review and edit or delete any of the utterances. You click Submit and Pega Voice AI re-builds the model with the new utterances. You have helped train the model on how to identify a customer request to change address. You save the changes.

You have reached the end of this video.