Training Pega Email Bot with Pega GenAI

U+ Bank wants to improve the text analysis performed by Pega Email Bot. You are asked to help train the email bot to identify customer emails that express anger and frustration. To support this scenario, you must update the training data for the MySupport channel.

You send an email to test how Pega Email Bot processes the message. The NLP model identifies that the test email has negative sentiment and the content is associated with the Angry topic.

You click the Training data tab, and then click Add records using GenAI. In the dialog, you select the topic Angry, and set the number of records to 5. You use the email content from the test email as sample content:

I am dissatisfied with the handling of my credit card issue. Despite reaching out to your support team multiple times, the problem with payments not being credited to my account remains unresolved.

You click Create record. Pega GenAI then uses the sample data to generate 5 similar training records and displays the records for your review. You review each record, check that it has identified the topic as Angry, and then select the records and click Reviewed. You rebuild the model, so that the new records are reflected in Pega Email Bot's learning.

This concludes the demonstration. You learned how to use Pega GenAI to improve the capability of an NLP model.