Here is a situation most data engineers have been in. A pipeline breaks, the error traces to a column transformation, and the business rule that governs that column lives in a Confluence page from two

Markdown supported · ⌘↵ to post
No responses yet.