MLOps is the practice of reliably building, deploying, and monitoring ML systems.

It combines software engineering, data engineering, model training, deployment, monitoring, governance, and retraining workflows.