Summary
Souvik Ghosh is a senior AI research and engineering leader with two decades of experience translating statistical rigor into large-scale product impact, currently serving as Sr. Principal Scientist in Mountain View. He has driven transformative AI work at LinkedIn—scaling model sizes by orders of magnitude to power personalization and leading a DMA compliance effort across 300 engineers to make 800+ ML models compliant. As a former Director of Engineering who led a 100+ person AI Foundations team, his work touches personalization, NLP, multimedia, federated learning and optimization across feed, ads, jobs and marketplaces, delivering measurable business wins like double-digit improvements in engagement and ads CTR. Trained as a PhD in Probability and Statistics from Cornell with prior academic and research roles at Columbia and Yahoo!, he blends deep theory (time series, extreme values) with production instincts. Notably, he also runs AI training initiatives internally—having built LinkedIn’s popular AI Academy that trained thousands—reflecting a rare mix of technical leadership, regulatory savvy, and talent development.
9 years of coding experience
17 years of employment as a software developer
Doctor of Philosophy (Ph.D.) Probability and Statistics, Doctor of Philosophy (Ph.D.) Probability and Statistics at Cornell University
Masters; Bachelors Statistics, Masters; Bachelors Statistics at Indian Statistical Institute, Kolkata
English, Bengali, Hindi