I can understand not being used to writing pure algorithms for an interview, so I guess that is something you can certainly spend a bit of time working on improving. But I’m a bit curious about being able to solve a question in “a couple of hours”. In my experience most of the interview questions I’ve seen are of the “I know how to do this and I can implement it, maybe needing a little bit at the end for some edge cases” or “I have no idea how to do this, I’m just going to brute force to have something and then spin my wheels for the remainder of the time”. I can see the time being useful for competitive programming, where you often need to put things together in nontrivial ways, but I find that most interview questions are fairly straightforward once you know what the algorithm or data structure is. Unless you’re rederiving it in the fly, which I have done sometimes, but it’s still not something that I would spend a couple hours on.