We're more focused on the data use case right now, but I could see it being useful for large code bases as well.