人工智能 (Réngōng Zhìnéng)

The program trains engineers who work with artificial intelligence at the device level, not the interface level. The difference is fundamental: today, anyone can use ready‑made neural network services, but only a few know how to design a system that learns from a specific company’s data, withstands the load, is explainable, and doesn’t break when the data changes. It is this shortage that hinders the field.

A graduate understands the mathematics underlying the models, knows how to build a data pipeline from source to production, train and evaluate a model, deploy it in a service, and be responsible for its behavior after launch. The program is deliberately built around engineering practice: the theory is provided in the amount necessary for independent problem‑solving, not for retelling.