EIT Digital student turns her PhD project into a business

Yuping Yan's relationship with EIT Digital has seen her go from enrolling as a master's student to working on her PhD, while also developing a cutting-edge machine learning business that ensures data privacy. She says her career has benefitted from EIT Digital's philosophy of combining top-level academic training with hands-on entrepreneurship experience.

The EIT Digital PhD programme's preparation for running an enterprise included practical business courses, according to Yan.

"This business experience actually helps a lot. It helps us to build business models and cases and it was useful when I started my enterprise," she says. "We also learn to present our case to other investors or to get funding, and that's very important."

The enterprise Yan is developing involves the brand-new field of federated learning, a technique for providing security and privacy during machine learning.

"Federated learning is a machine learning approach that enables training models across multiple decentralised devices or systems while keeping the data local," according to Yan's paper describing the project. In a federated learning system, the data is not pulled onto one central server. Instead, the machine that is learning has individual sessions with separate data sets, thereby ensuring that the data stays private.

Yan's business, called FedlabX, makes a unique GDPR compliant machine learning platform available for other companies to use.