Any recommendations for using Matillion to build solid data checks (DataOps)?

As a ETL developer, we are often struggling with seeing how are changes are effecting the whole pipeline, and many times we send bad data that goes unnoticed for some time.

 

Even delta counts aren't always very helpful, but that's basically all we got.

 

It would be great if we could capture key data elements that would help us identify issues as we are unit testing.

 

Just wondering if any of you have built something in matillion that helps you reduce errors (shared jobs, etc)?