This is probably a non-standard case study. We have actually been bug hunting for around the last two months, e.g., small issues that are not showstoppers. They are more annoying. We have been going through and squashing them. There was one bug where we couldn't find a specific reason as to why it occurs randomly through the company. We know why it occurs now. We don't really know why it pops up randomly, but all the printers will be gone. It's like, "Okay, what?" So, the traditional solution was to restart the workspace, but this was happening two to three times a week or even sometimes multiple times a day. It's like, "Well, this is annoying." So, we went on a bug hunt. We opened up a ticket and got engineers involved, pulling more resources into it. We had conference calls on debugging solutions. That's what I call a deep dive. It was to the point where we were on a conference call with engineers stacked on top of engineers who were all yelling at me to get logs as well as pull this and do that. Eventually, it just turned out to be a Group Policy Object (GPO). Something that should have been done during the initial setup and configuration almost three years ago wasn't done and that represented as a bug where you lost printers. While I could have dealt with it, it wasn't showstopping. It was annoying for me because I would get the call regardless if I was here or not. The ability to pull talent together from CDW, Amazon, and Microsoft is something that I could never do because I'm a lowly peon. Unless I am putting down seven or eight figures, Microsoft won't deal with me. However, CDW has the ability, through their team of talent, to say, "Hey, can you jump on this call?" and a very bored engineer is like, "That is a Group Policy Object. Here is the file to fix it." This is a very long-winded story about how I can lean on the CDW cloud teams (their Managed Cloud Services) in a way that Google doesn't really help me. CDW Managed Cloud Services has found pitfalls and various things in the infrastructure, e.g., in routes, IP ranges, and security loopholes, that didn't need to be there. They stripped it down and went to a very flat topography, making it less complicated. So, there are less things to trip over and go wrong, which means less downtime.
