I was in an infrastructure review with the team the other day and noticed something interesting. Everyone was comfortable talking about compute optimization. Rightsizing instances? Sure. Tweaking autoscaling? No big deal. We even talked about moving workloads around and nobody seemed worried. But as soon as someone mentioned cleaning up unused storage in production, the whole conversation changed. Nobody disagreed that we were wasting money. We all knew there was storage sitting there that probably didn't need to be. The problem was that nobody wanted to be the person to touch it. Maybe it's because storage feels a lot more permanent. If you mess up compute, you can usually roll things back pretty quickly. With storage, people immediately start thinking about deleted data, broken applications, and late-night calls.…