Tianyuan Yao is a Senior Machine Learning Scientist at Pangea, focusing on the development of peer-to-peer federated learning systems with homomorphic encryption. He also contributes to medical imaging research at Pangea.
Tianyuan works on an NIH-funded project focused on developing privacy-preserving artificial intelligence systems for healthcare applications. His work centers on designing peer-to-peer federated learning architectures that allow multiple healthcare organizations to collaboratively train machine learning models while keeping sensitive data securely within each institution. By combining federated learning with advanced encryption techniques, the project aims to enable secure, scalable, and trustworthy AI systems for analyzing distributed clinical data.
He is also involved in medical imaging research at Pangea, where his work focuses on applying machine learning methods to complex medical imaging data. His research interests include representation learning for heterogeneous and multi-modality imaging data, as well as adaptive modeling approaches for advanced imaging analysis, contributing to the development of AI-driven tools that support improved analysis and interpretation of medical imaging data in healthcare research.