Production Governance and MLOps on Databricks

Overview

Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off

One annual plan covers every course and certificate on Coursera. 40% off for a limited time.

This intermediate course provides a practical, hands-on exploration of Databricks Governance, focusing on the essential tools and workflows for managing and securing your data lakehouse. You will learn to navigate and control access to your data assets using Unity Catalog, the foundation of Databricks governance. The course covers the core hierarchy of metastores, catalogs, schemas, and tables, and teaches you how to manage them programmatically using the Databricks Python SDK, CLI, and VS Code extension. Beyond foundational access control, you will master the skills to implement modern CI/CD and MLOps practices directly within the Databricks environment. You'll learn to integrate Databricks Repos with GitHub, automate notebook testing and deployment with GitHub Actions, and understand the architectural considerations for managing machine learning models in production. Finally, you will explore how to ensure ongoing data reliability by setting up and understanding Lakehouse Monitoring for data quality and freshness. This course is unique because it moves beyond theory, demonstrating how to apply these governance concepts with the actual tools and code used by data professionals. By the end, you'll be equipped to build, deploy, and monitor secure and reliable data pipelines and AI applications on the Databricks platform

Syllabus

Unity Catalog Governance

This module establishes the foundation of Databricks governance through Unity Catalog. You'll navigate the metastore-catalog-schema- table hierarchy, set up role-based access control using service principals and GRANT/REVOKE statements, and learn to manage your governance setup programmatically with the Databricks Python SDK, CLI, and VS Code extension.

CI/CD and MLOps

This module covers the workflows that take Databricks code from a developer's laptop to production. You'll integrate Databricks Repos with GitHub using branching strategies and code review, automate notebook testing and deployment with GitHub Actions, and build a complete MLOps pipeline that serves a GenAI application through a model serving endpoint.

Monitoring and quality

This module closes the production loop with Lakehouse Monitoring. You'll enable quality and freshness monitoring on Unity Catalog tables, interpret monitoring results to detect data anomalies and drift, and review the recommendations that turn a working pipeline into a production-ready governance setup.