Azure Machine Learning pipelines are workflows of machine learning tasks that can be automated. These pipelines can be configured with triggers to run automatically in response to specific events. Triggers can be schedule-based (e.g., run daily) or data-based (e.g., run when new data is detected in a datastore). This functionality is ideal for implementing MLOps practices, such as automated model retraining, ensuring the model stays current without manual intervention.

B. Data labeling is a feature for annotating data to create training sets, not for automating retraining workflows.

C. A compute instance is a managed cloud-based development environment, not an automation or triggering mechanism.

D. Endpoint scaling adjusts the compute resources for a deployed model to manage inference load, not for retraining.

1. Microsoft. (2023). Trigger the run of a machine learning pipeline from a Logic App. Microsoft Azure Documentation. Retrieved from https://learn.microsoft.com/en-us/azure/machine-learning/how-to-trigger-pipeline-logic-app. (This guide shows how to create triggers for pipelines.)

2. Microsoft. (2023). Create and run machine learning pipelines by using the Azure Machine Learning SDK. Microsoft Azure Documentation. Retrieved from https://learn.microsoft.com/en-us/azure/machine-learning/how-to-create-machine-learning-pipelines. (Under the "Schedule a pipeline" section, it details how to set up schedule-based triggers for automated runs.)