Skip to main content

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

Configuring text prediction

Updated on May 10, 2021

Define how Pega Platform natural language processing (NLP) analyzes conversations on your Voice AI channel by configuring the behavior of the text analyzer that is associated with the channel.

Targeted text prediction settings generate more accurate suggestions for sales representatives and help them deliver an improved customer experience.
Pega Sales Automation Implementation Guide
  1. In the navigation pane of App Studio, click Channels.
  2. In the Current channel interfaces section, select the Voice AI channel interface for which you want to configure text analyzer settings.
  3. On the channel configuration page, select the Behavior tab.
  4. Optional: To modify the list of topics that make up your overall business logic, click Edit topics, and then add or remove topics as needed.
  5. Optional: To start collecting data from live interactions and automatically add records to the Training data tab when the NLP engine fails to uniquely detect a topic, select the Enable model training checkbox.
  6. To ensure that the NLP engine does not fail to detect entities based on short utterances in a conversation, edit the default language setting of the text analyzer:
    1. Click Open text prediction.
    2. Select the Settings tab.
    3. In the Language settings section, in the Select language list, select English.
      Note: Currently, Voice AI supports only English.
    4. In the Language detection field, select Specify language, and in the Language list, select English.
  7. Save your updates to the channel configuration by clicking Save.

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