A company built a GPT assistant trained on supply chain research and their own data to evaluate orders and flag issues. It cut manual order handling from 80-90% down to 5-10%, but now risks losing experienced staff who traditionally did this work.
And what they did was they said, okay, we've got to teach a GPT everything about supply chain. So they fed it all the college literature on supply chain, and now they had an expert on supply chain. Then they added a layer saying, okay, what are we as a company doing? So you give it an SME that feeds it what, what, what are we doing here at the company for, for supply chain. And then it feeds its input from our system to show what are we actually doing. And so now you can take a new person and you have this assistant next to them, and they would be able— the system will be able to tell you, hey, here is what's good, here's what's bad, here's how you do this, here's how you do that. And they were able to— so they had orders where they had to touch like 80, 90% of the orders, and now they're down to like 5 to 10% of the orders they have to touch because this, this assistant is now taking care of categorizing the orders and solving some of the problems already. And so that they can process these orders much faster. But they, they did it basically, uh, around the fact that they're losing that core talent that has just been doing this for years and years and not complaining about all that extra thing they got to do to get an order through, right?
