Why do we need to go back to CDC for automated test cases when we already have curated data loaded into the curated zone?

Automated test cases often need to verify that the data transformation pipeline produces the expected results for new or changed data. The curated zone contains the final, cleaned data, but it may not reflect the incremental changes that occur in the source. By re‑reading the CDC stream, you can generate test inputs that represent the exact changes that the pipeline will process. This allows the tests to run against realistic, up‑to‑date data without re‑ingesting the entire dataset. It also ensures that any new mapping logic or business rules are validated against the current source state. In short, CDC provides a lightweight, incremental source of truth that keeps automated tests fast, accurate, and aligned with the live data.