The Data Engineering Paradox: Why We're Still Reinventing the Wheel

What a great read. Thanks for putting in the time and effort Maxime. I was wondering if a declarative data stack would help. Isn't a declarative approach the solution to much repetition? Much like in Kubernetes, you define what you want, and implementation code is abstracted away and not duplicated. The declarative approach lets you automate, version, and all the good things of SW's best practices.

Much of it also reminds me of the days of Data Warehouse Automation, when I could template all of the tables with 2-3 templates and generate the whole DWH. Actually, these tools are still out there and used daily. Time to revive?