Skip to main content

         This documentation site is for previous versions. Visit our new documentation site for current releases.      

Configuring case suggestions

Updated on December 3, 2021

In your Voice AI channel configuration, create case suggestions so that Voice AI can suggest appropriate cases to CSRs. Case suggestions prevent the need for CSRs to manually find case types that are relevant to a customer inquiry.

Pega Customer Service Implementation Guide

Creating case suggestions involves adding one or more case types that you define in your application to the Voice AI channel configuration. When you add a case type, you specify a command, which is a topic that triggers Voice AI to suggest that case type to the CSR if Pega Platform natural language processing (NLP) detects that command. After the CSR launches the case, Pega Platform NLP also detects entities that the customer mentions, and then automatically propagates them to the case type fields that you map.

To quickly start using case suggestions without training your Voice AI model, you can directly add phrases in the Text analysis tab while configuring a case suggestion.

Note: Voice AI can only suggest case types that you specify in the Cases and Data tab in your Pega Customer Service application.

  1. In the navigation pane of App Studio, click Channels.
  2. In the Current channel interfaces section, select the Voice AI channel interface to which you want to add the case suggestion.
  3. On the Configuration tab, in the Suggested cases section, click Add case type.
  4. In the Create command dialog box, on the Response tab, in the Case type list, select a case type that you want Voice AI to suggest to your CSRs.
  5. In the Create case command field, define the command that you want to use:
    • To use an existing command, in the list of topics, select an item.
    • To create a new command, enter a topic that you want to use.
    Result: If you add a new command, the command is also added to the list of topics that you can select when you configure knowledge suggestions, and to the list that opens when you click Edit topics in the Text Analyzer section of the Behavior tab of your channel. For more information, see Configuring text prediction.
  6. On the Text analysis tab, add comma-separated lists of words to trigger a suggestion for the case type:
    1. In the Approximate match field, enter the words that can trigger the case suggestion but that are not required for the case suggestion to be triggered.
    2. In the Must match field, enter the words that must be present in the conversation to trigger the case suggestion.
    3. In the Never match field, enter the words that prevent the case suggestion.
  7. On the Entities extraction tab, define the autofill settings for the case type by mapping the entities that Pega Platform NLP detects in the conversation to the Pega Customer Service case properties:
    1. Click Add mapping.
    2. In the Entity list, select an entity.
    3. In the Case property list, select the field to which you want to map the entity.
    For example: If you want the country code to populate when CSRs select a country, select Country as the entity, and map it to the Country code case property.
  8. Click Submit.
  9. Optional: To add additional case suggestions to meet your business requirements, repeat steps 3 through 8.
  10. Save your updates to the channel configuration by clicking Save.
Result: The Suggested cases section on the Configuration tab of your Voice AI channel displays the new case type. When Pega Platform NLP detects a command or text in a conversation, Voice AI suggests the associated case type to the CSR.


Pega Customer Service 8.7 Pega Customer Service for Communications 8.7 Pega Customer Service for Financial Services 8.7 Pega Customer Service for Healthcare 8.7 Pega Customer Service for Insurance 8.7

Have a question? Get answers now.

Visit the Support Center to ask questions, engage in discussions, share ideas, and help others.

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best. is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us