Instances & Endpoints

To deploy and manage the Agent Coach model effectively, you must create the Agent Coach instance comprising three essential components:

• Model Training Service URL—This core component enables you to configure and execute the training process, thus facilitating model training by providing necessary data and creating an optimal training environment.

• Recommendations Service URL (Pre-Production and Production)—These endpoints are essential for providing recommendations based on the trained model. The Pre-Production environment is used for testing, and Production is the live production environment where real-time recommendations are made to users.

• Logging Service URL (Pre-Production and Production)—These endpoints are required to monitor the performance of your model. Both Pre-Production and Production logging endpoints enable you to track the training progress, identify issues, and ensure that the model functions optimally in various environments.

Once you define these instances and endpoints, you can initiate the Agent Coach model training by creating a Job ID.

Note: The Job ID serves as a unique identifier for each training process, facilitating management and monitoring.