I like a code review process better than a coding exercise for sure, but it’s not clear to me that it helps avoid LLM-enabled interview “performance enhancement

I like a code review process better than a coding exercise for sure, but it’s not clear to me that it helps avoid LLM-enabled interview “performance enhancement”. Avoiding “take home” is a big step, but even in real-time determined LLM use can make someone look better in an artificial context than they will be in practice. And, as others have pointed out being good at code reviews is one facet of development, and over representation of it at the detriment of other skills isn’t great. If your code review conversation spans all those areas, maybe that’s the best we can do? I don’t have a great alternative (I’ve used a similar approach for years), and the app is certainly interesting. I might be giving it a try. Yes, using LLMs is a skills too.