It's pretty naive to think you can just hire random people to step in and support complex business systems part-time. Sure, you can have a global ops team that

It's pretty naive to think you can just hire random people to step in and support complex business systems part-time. Sure, you can have a global ops team that can deal with basic system issues like "this machine ran out of disk space", but not "we're seeing a spike in 500s in our payment system because something that got deployed earlier today broke a workflow". And saying "well just roll back!" isn't a solution either. Maybe you can't roll back. Maybe that won't fix it because data is now corrupt. Maybe it wasn't even a deploy to that system that broke things, but some feature flag in some distant system. No one except the team itself can diagnose these issues appropriately.