About

Garrett Goh is an accomplished ML applied scientist engineering manager with over 10 years of relevant experience in the field of AI and deep learning. He currently works at Microsoft, Xbox, where he leads and manages an applied ML team that drives a 10-fold increase in custom ML solutions for content, including images, videos, audio, and text. His team's models are up to 200% more efficient than human-level baseline and are currently deployed across several product lines, including all major IP franchises. Garrett is a subject matter expert in machine learning and has been appointed to the Microsoft Worldwide Machine Learning SME panel. He has published work at top-tier machine learning conferences and workshops, including NeurIPS and KDD, and has presented at industry conferences such as O'Reilly AI Conference. His work has garnered over 1000 citations, with an h-index of 15. Prior to joining Microsoft, Garrett worked as a senior research scientist at Pacific Northwest National Laboratory (PNNL), where he led a research team that pioneered one of the first self-supervised learning methods that incorporates domain knowledge for computer vision CNN models, NLP/NLU-styled LSTM models, and multimodal models. He managed a 3-year $1 million research grant and supervised two students on applied deep learning research projects, resulting in seven publications in peer-reviewed machine learning conferences and workshops, including NIPS, KDD, and IJCAI. He was awarded the early career distinguished Pauling Fellow award in 2015, with a selectivity rate of only 1% and two awardees company-wide. Garrett holds a Doctor of Philosophy (Ph.D.) degree from the University of Michigan, where he specialized in AI and deep learning. He has also worked as an adjunct lecturer at Harvard University, co-lecturing a graduate-level course on modern deep learning in computer vision and practical applications. Garrett's tech stack includes research scientist, ML, senior software engineer, computer vision, AI, deep learning, supervised learning, image processing, NLU, and NLP. He is a highly motivated and results-driven individual with a proven track record of transforming ambiguous business challenges into tangible applied ML solutions to enhance the business vertical.