Here's a pattern I keep running into with startups doing AI training on Azure: they get H100 capacity approved, spin up VMs, launch a multi-node training run, and watch it crawl. Or worse, NCCL hangs...

Most startups I work with have Service Health alerts configured on at least one subscription. That part's fine. The problem is they usually have it set up for Service Issues only, on their main subsc...