I have a pet theory that any given system has a maximum cleverness budget. Make sure to spend it on the things which actually matter in your problem domain.

Interesting data products often have to spend a lot of that cleverness budget in the algorithmic layer – machine learning and statistics. So when I work on that kind of thing, I've learned to have a real appreciation for boring data storage. I'll even happily use MySQL; the ways in which MySQL can suck are very well-understood and that's a hugely underappreciated virtue.