What you'll learn

Machine learning models are only as effective as the systems that support them. In this course, Introduction to MLOps, you’ll learn how to build a streamlined, end-to-end MLOps pipeline using a minimal set of open-source, Linux-friendly tools. First, you’ll explore how model registries and feature stores work together to promote collaboration, versioning, and repeatability, using MLflow and Feast to register, track, and manage your models and features throughout their lifecycle. Next, you’ll learn how to apply version control techniques to both data and models, using Git and DVC to ensure that experiments remain reproducible and that your pipeline responds automatically to changes in code or training data. Finally, you’ll discover how to monitor model performance and deploy it across hardware accelerators, leveraging tools like Evidently for drift detection and BentoML with Docker for scalable, optimized inference. When you’re finished with this course, you’ll understand how to implement robust MLOps workflows using open-source tools, maintain model quality over time, and deploy performant machine learning services with confidence.