Testing and CI/CD for Data Pipelines

<p>Follow a data engineer through a real incident where correct code still produced wrong numbers, then build the testing and deployment safety net that prevents it from happening again: unit tests for transformation logic, integration tests against a realistic dataset, a GitHub Actions CI pipeline that runs on every pull request, and a staged, blue/green-style rollout with a rollback plan for schema changes.</p>

Tools: Python, pytest, dbt, GitHub Actions, DuckDB, PostgreSQL