Machine Learning Ops (MLOps) platforms operationalize traditional ML models from training through production deployment, serving, monitoring, and retraining. Core capabilities include experiment tracking (hyperparameters, metrics, artifacts), model registry and versioning, feature stores, training-pipeline orchestration, model deployment to batch/online/edge serving targets, model monitoring (data drift, concept drift, accuracy degradation), A/B model rollout, and governance/lineage for regulated ML. The category spans experiment-tracking tools (MLflow, Weights & Biases, Comet, Neptune.ai, ClearML), end-to-end ML platforms (Databricks ML, AWS SageMaker, Google Vertex AI, Azure ML), and specialized model-serving products (Seldon, BentoML, Modal).