Serving:

Using a deployed machine learning model for real-time inference or predictions in production systems. It involves hosting the model on a server or infrastructure that can receive input data, process it using the model, and provide a corresponding prediction.