I don't think data engineering is hard or needs much resources. It does need people who care and who will think about things.
Anyone who comes at a problem with the mindset 'this is going to he hard' probably lacks experience and will throw big-data frameworks at it, really screwing things up. The most significant, and valuable, resource needed is thought first, and knowledge+experience second.
I always think that database design is pretty simple, but having worked on numerous databases designed by other, some people get it very badly wrong. "Wrong" maybe isn't the correct word, because you can use the database for what it is needed but it just requires extra joins and hacks at the application level to get around poor design.
Depends on context, you're talking in general, but in particular as the GP post said, there can be pressure from management to rush things, different data formats, bad or missing data and so many other pitfalls. It's hard when it's hard.
You can write a simple test to check a function is working correctly, but how do you make sure your 100,000 item database doesn't have corrupted or missing data caused by the latest pipeline update, especially if the corrupted parts are rare?
Anyone who comes at a problem with the mindset 'this is going to he hard' probably lacks experience and will throw big-data frameworks at it, really screwing things up. The most significant, and valuable, resource needed is thought first, and knowledge+experience second.
All IMO anyway.