Viet Duong is a researcher in the Data Science & System Security department at NEC Laboratories America, based in Princeton, New Jersey. He holds dual bachelor’s degrees in data science and mathematics, as well as an M.S. in computer science, all from the University of Rochester. Dr. Duong earned his Ph.D. in Computer Science from the College of William & Mary, where he was advised by Dr. Huajie Shao and received the Stephen K. Park Graduate Research Award in 2025. His doctoral training at the intersection of machine learning, weak supervision, and interpretability forms the foundation of his applied research at NEC.

At NEC Laboratories America, his research centers on building reliable and interpretable machine learning systems that can operate effectively in real-world, high-stakes environments. A central focus of his work is developing concept-based frameworks that make model reasoning transparent and understandable to users across varying levels of domain expertise, while reducing the need for costly annotation and supervision by domain experts. His approach to interpretability goes beyond post hoc explanations by embedding human-aligned concepts directly into model architectures, enabling meaningful and trustworthy decision support.

He also investigates out-of-distribution detection across modalities, including code, text, and multimodal data streams, a capability critical for deploying AI systems in settings where inputs may deviate unpredictably from training conditions. His work has been recognized with a Best Paper Award at the ACM SIGKDD Conference on Knowledge Discovery and Data Mining and published in venues including ICSE and Transactions on Machine Learning Research. By combining principled machine learning theory with practical system design, his contributions advance the responsible deployment of AI across enterprise and safety-critical domains.