Agent Coach Management

Agent Coach comprises a comprehensive set of APIs with multiple endpoints, all hosted in the Azure datacenter. Customers have the flexibility to access at least one instance, with the option to configure multiple instances as needed. This capability facilitates the training of multiple lines of business (LOB) using the Agent Coach model.

Note: Nuance recommends that customers deploy only one instance of Agent Coach, which is trained to handle multiple lines of business (LOB).

Under Agent Coach Management, agents can select and modify top recommendations suggested by the recommendation system. These recommendations are created by analyzing the current chat history and previous transcripts, leveraging prior responses to provide the most effective response suggestions.

The suggestions are context-driven, providing agents with timely recommendations based on the context. Agents maintain full control and can choose to respond directly with the suggested reply, make edits, or create their own responses.

To ensure that Agent Coach runs smoothly, you must complete the following tasks:

• Setting up an instance—It includes connecting NDEP to the Agent Coach APIs, including runtime, training, and logs. For more details, refer to the Instances & Endpoints section.

• Training models—It includes training AI models using specific data sets, such as Business Units (BU) and Agent Groups (AG). For more details, refer to the Model Training section.

• Curation process—It includes validating the response pool of a trained model. For more details, refer to the Recommendations Curation section.

• Enabling Agent Coach for different Sites or Agent Groups—Ensure that Agent Coach is properly set up and available at each given location. For more details, refer to the Agent Group Enablement section.