If you’ve ever been frustrated by a test database filled with nonsense, duplicates, sensitive information, or YOLO-style restored production backups, this workshop is for you.

Join Peter Kruis and Tonie Huizer for a practical, SQL Server focused workshop on building test databases that are safe, realistic, refreshable, and actually useful for development and testing.

We will cover the most common challenges teams run into across different roles, including data masking and privacy, balancing database size versus performance, handling application settings, supporting multiple coworkers working in parallel, and managing reset, refresh, and versioning of test environments.

From there, we will show how to choose the right dataset for the job: starting from an empty database with seed data, using masked production-like data, or working with a representative subset of production. We will then turn that into reusable golden images and show how teams can spin up copies for individual development and test tasks.

A key part of the workshop is automation. We will demonstrate how database provisioning can be triggered by the team’s workflow in Azure DevOps rather than by someone manually running local scripts. Using practical examples in PowerShell and YAML, we will show how fresh databases can be provisioned automatically when new work starts, how work-in-progress environments can be stashed and resumed when switching context, how specialized pipelines can support different database needs, and how stale environments can be cleaned up as part of the normal flow.

The workshop focuses on SQL Server, real-world examples, and demonstrations. While the implementation shown is SQL Server based, the underlying ideas and patterns can also be applied to other relational database platforms.