Operationalize machine learning models (MLOps)
Microsoft Learn
Last updated October 6, 2026
Operationalize machine learning models (MLOps) is a free, seven-module Microsoft Learn path that takes a model from first experiment to monitored deployment using Azure Machine Learning and GitHub. Machine learning operations (MLOps) applies engineering discipline to a model's lifecycle, so that training, release and monitoring can be repeated reliably. The first stage designs a training solution, compares models with automated machine learning, MLflow and the Responsible AI dashboard, converts notebooks into tracked training scripts and tunes hyperparameters. The second builds reusable pipelines, plans an MLOps solution, automates training with GitHub Actions, and registers, deploys, monitors and rolls back models through protected environments.
What you'll learn
- Framing a machine learning problem and choosing Azure Machine Learning services and compute for training
- Comparing models with automated machine learning, MLflow and the Responsible AI dashboard
- Converting notebooks into training scripts, running command jobs and tracking models with MLflow
- Tuning hyperparameters with sweep jobs
- Automating multistep workflows with reusable components and Azure Machine Learning pipelines
- Automating and validating model training with GitHub Actions
- Registering, deploying, monitoring and rolling back models with GitHub Actions and protected environments
Frequently asked questions about Operationalize machine learning models (MLOps)
Who is Operationalize machine learning models (MLOps) for?
Data scientists and machine learning practitioners who train models in Python or R and want to automate, deploy and monitor them on Azure Machine Learning with GitHub.
Is Operationalize machine learning models (MLOps) free?
Yes — Operationalize machine learning models (MLOps) is completely free to take.
What are the prerequisites for Operationalize machine learning models (MLOps)?
Programming experience with Python or R, experience developing and training machine learning models, and familiarity with basic Azure Machine Learning concepts.
Do you need to code for Operationalize machine learning models (MLOps)?
Yes — Operationalize machine learning models (MLOps) involves hands-on coding.
Why we suggest this course
For a data scientist who can train a model and now has to ship it and keep it working, this path runs from training experiments through to automated deployment and monitoring. Its order follows a real project: experiment, track, tune, automate, deploy and monitor. The final module includes rolling a model back, which is the step that matters when a release goes wrong. One thing to know: it is built on Azure Machine Learning throughout, with GitHub Actions in the final modules; the practices transfer to other platforms, but the services and commands do not.