Using the conversational channel

Enhance chat interactions among Pega Intelligent Virtual Assistant (IVA), users, and your application by using the conversational channel. For the IVA to correctly recognize and intelligently respond to user commands, classify and add training data to the text analytics model.

Before you begin: To use a conversational channel, for example, IVA for Unified Messaging, first build the IVA by defining its behavior and simulating a chat conversation. For more information, see Building a conversational channel.

When you train a text analytics model for an IVA, the system learns to correctly respond to a user message based on the recorded data that you update. The system learns to detect the correct subject matter (topic) and phrases (entities) in the message, such as names, dates, or postal codes. For example, when a user chats with an IVA to request a bank loan, the system intelligently detects the topic and entities, and automatically starts a bank loan business case. In this business case, the user is then asked a series of follow-up questions about their personal information, income, and financial history.

To learn how to record training data, create training data, and then apply changes to the text analytics model for the IVA, see Training the model for the IVA channel.