Summary
Yuan Shangguan is a Staff Software Engineer in the San Francisco Bay Area with ~12 years of experience applying machine learning and speech research to production devices. He has led on-device speech and compression work at Google and Meta’s Reality Labs, shipping voice features for Quest, Portal, and smart glasses, and now focuses on making LLM tool agents useful and safe at Google DeepMind. Yuan’s background spans academia and industry—from multimodal emotion classification research to patents in adaptive decoder beam pruning—showing a blend of theory and deployable systems. He excels at squeezing state-of-the-art models into constrained hardware through neural architecture search and model compression. Colleagues rely on him to translate cutting-edge research into robust, product-ready AI for edge devices.
12 years of coding experience
9 years of employment as a software developer
Master's degree Computer Engineering Artificial Intelligence, Master's degree Computer Engineering Artificial Intelligence at University of Michigan
Bachelor of Engineering (B.Eng.) Computer Software Engineering, Bachelor of Engineering (B.Eng.) Computer Software Engineering at Dartmouth College
A levels Science Stream, A levels Science Stream at Raffles Institution
German, Chinese, Chinese