The problem
The shop knew it was busy but not what was actually making jobs late. An open machine doesn't help if the qualified operator or the material isn't there, and a resource that looks overloaded isn't always the one holding up shipments.
What they did
They connected Skody to ProShop so it builds a finite schedule from live jobs, machines, operator skills and material dates. In weekly scheduling meetings they run 'what if' versions: give an office employee mill skills on paper and re-run the plan, move a material date earlier, mark someone absent, or add a new order, then check whether customer ship dates actually improve.
Reported results
Skody's summary page says on-time delivery went from under 80% to over 90% within two months, and the case study says revenue has doubled since they started using it. The shop's president says a new plan after an absence can be checked in a couple of minutes, and that busy weeks mean 8-10 scenarios in an hour-long meeting. These are vendor-reported numbers.
Our take
The useful idea here isn't the AI, it's the habit: before you hire, buy a machine or expedite material, check whether that change would actually move ship dates. Often the 'obvious' bottleneck isn't the one that matters. Keep in mind this is the vendor's own write-up, the revenue growth has many causes, and machine-plus-people scheduling sits in Skody's custom-priced tier, above the $1,500/month starting plan.