I'm doing some leetcode problems and finding it very disconnected from the likely jobs I'll have in the future. I actually have a lot of fun with it since I've been out of school for 15 years now and it really gets me thinking. But it's also frustrating because I know with about a 100% chance that these skills won't be used in a job. I'm quite certain that my success in a future role will be dependent on my ability to learn the narrow domain problems the company will solve.

So I both 1) want to be hireable - gotta know my data structures and algorithms, I get it; and 2) have more skill in solving real-world problems.

But grinding leetcode doesn't make you a better software engineer (right?)

Previously I've never had a problem with this, as my side projects have all been somewhat novel and required me to both work through interesting algorithms and also build real software to be used by real people.

But now it seems nearly all interesting algorithms or problems to solve have been packaged into a nice library to use. And a real business would want you to use that because it will be less error-prone and faster to implement. But I like solving the fun leetcode problems, I just feel like I'm prioritizing for the interview and not actually being a better developer.

What do I do so that I avoid just gluing things together for the rest of my life?