Summary
Shiyang Chen is a Staff Data Scientist at Meta with nine years of experience building data-driven models and deploying advanced ML solutions for ads, GenAI, and business products. He holds a PhD in Mechanical Engineering and dual master’s degrees in applied mathematics and naval architecture, bringing deep theoretical grounding in stochastic processes, numerical linear algebra, and convex optimization to practical problems. At Meta he now contributes to Llama training data and generative AI efforts after leading analytics and modeling work for Facebook Ads; prior roles include time-series modeling for customer behavior at Citi and research on stochastic differential equation models at the University of Michigan. He codes fluently in Python, R, MATLAB, C/C++ and has applied neural networks, gradient boosting, SVMs and probabilistic inference to real-world systems. Known for extracting novel features from time series and linking eigenvalue behavior to system stability, he blends academic rigor with product impact to turn complex mathematical insights into scalable solutions. Based in New York, he combines research pedigree with hands-on production experience across large-scale data platforms.
9 years of coding experience
10 years of employment as a software developer
Master of Science (MS) Naval Architecture and Marine Engineering, Master of Science (MS) Naval Architecture and Marine Engineering at University of Michigan
Doctor of Philosophy (Ph.D.) Mechanical Engineering, Doctor of Philosophy (Ph.D.) Mechanical Engineering at University of Michigan College of Engineering
Master of Arts (M.A.) Applied Mathematics, Master of Arts (M.A.) Applied Mathematics at University of Michigan College of Literature, Science, and the Arts
Bachelor of Science (BS) Ocean Engineering, Bachelor of Science (BS) Ocean Engineering at Shanghai Jiao Tong University
Chinese, English