Alexander Ratner
About Alexander Ratner
Alexander Ratner is the Co-Founder and CEO of Snorkel AI and an Affiliate Assistant Professor at the University of Washington, both roles he has held since 2019. He has a background in computer science, holding a PhD from Stanford University, and has contributed significantly to the field of weak supervision and data-centric AI through various research initiatives.
Work at Snorkel AI
Alexander Ratner serves as the Co-Founder and CEO of Snorkel AI, a position he has held since 2019. Under his leadership, the company focuses on advancing weak supervision and data-centric AI methodologies. Ratner has been instrumental in developing Snorkel DryBell, a case study that showcases the deployment of weak supervision at an industrial scale. His work has contributed to the company's reputation in the AI field, particularly in creating classifiers that perform comparably to those trained with large hand-labeled datasets.
Education and Expertise
Ratner holds a Doctor of Philosophy (PhD) in Computer Science from Stanford University, where he studied from 2014 to 2019. He also earned an A.B. in Honors Physics from Harvard University, completing his studies there from 2007 to 2011. Additionally, he attended The Lawrenceville School from 2003 to 2007. His educational background equips him with a strong foundation in both physics and computer science, which informs his research and professional endeavors in AI.
Background
Before co-founding Snorkel AI, Ratner gained experience as a PhD student at Stanford University from 2014 to 2019. He also interned at Konarka Technologies in 2010, where he worked in the Physics Group for three months. In addition to his work at Snorkel AI, he has been an Affiliate Assistant Professor at the University of Washington since 2019, where he continues to contribute to academic research and education.
Achievements
Ratner has authored several research papers on weak supervision and data-centric AI, with publications in prominent conferences such as NeurIPS and ICML. He has developed frameworks like Snorkel MeTaL for weak supervision in multi-task learning, which was published at SIGMOD. His collaborative efforts with organizations such as the US FDA and Veterans Affairs have resulted in significant improvements in predictive performance, demonstrating the practical impact of his research.