The way I did it then probably wouldn't work now. I literally just wrote a perl script to automate pushing copies out to all the servers. They were doing all by hand before I got there.

If I had to do it today, I'd be looking at GitHub actions or some other similar CI/CD tool. If you're already using those, then I'd ask what is the slowest part of the process, and how can we speed it up?

To answer that question you obviously first need telemetry and monitoring to track each step of deployment. That telemetry will be useful in not only finding slow spots but tracking your improvement, which will help you both in operating as well as justifying your work to others.