[D] has anyone ever worked on a machine learning model for "queues"?

Has anyone ever worked on a machine learning model for "queues"? Suppose there is a bakery: the bakery has has "n" people working, "m" people in line" and "q" orders that they are currently working on. The bakery is interested in making a machine learning model that predicts how long a customer will have to wait before the customer's order is ready and how long will the next customer have to wait before they can place an order. Has anyone ever come across a machine learning model which can predict waiting and processing times? I have seen examples online where people try fitting exponential distributions to historical waiting times and see how well they fit, as well as trying different m/m/k combinations... but has anyone ever come across an instance where machine learning algorithms (e.g.