Deployment & Scaling · Advanced
Azure Machine Learning: When You Still Need a Full ML Platform
Foundry did not replace Azure ML. Learn where custom models, training pipelines, MLflow tracking and managed endpoints still matter - and how to avoid its expensive traps.
- 55 min lesson
- intermediate level
- Azure AI Track
- Narrated presentation
What you’ll learn
- Azure ML remains correct for tabular, forecasting, anomaly detection and custom classifiers
- Workspaces, datastores, compute instances versus clusters, managed endpoints
- Jobs and pipelines as code; notebooks for exploration only
- MLflow as the native tracking and model registry interface
- Online endpoints with blue-green traffic split; batch endpoints for scheduled scoring
- Idle shutdown on compute instances and scale-to-zero clusters as cost discipline
- Forgotten online endpoints bill continuously with zero traffic
Inside this lesson
A narrated, slide-by-slide presentation with 1 prediction exercise, 1 knowledge check, 3 videos. You answer as you go, and the platform checks your work.