Jelani Nelson
Theoretical Computer Scientist, Anthropic; Professor, UC Berkeley (on leave)
About
Jelani Nelson is a theoretical computer scientist known for his work on streaming and sketching algorithms — methods that process massive datasets using very little memory. He earned his BS, MEng, and PhD at MIT, taught at Harvard from 2013 to 2019, and then became a Professor of Electrical Engineering and Computer Sciences at UC Berkeley, where he served as Chair of the Computer Science Division. In 2011 he founded AddisCoder, a summer program teaching algorithms to Ethiopian high school students that has become a remarkable talent pipeline to institutions like MIT, Harvard, and Stanford. In July 2026 he took a leave of absence from Berkeley to join Anthropic as a theoretical computer scientist. His honors include the ACM Eugene L. Lawler Award and the Presidential Early Career Award for Scientists and Engineers.
Key Contributions
- Developed foundational results in streaming and sketching algorithms, including optimal bounds for dimensionality reduction and the count-distinct problem
- Founded AddisCoder (2011), a free summer coding academy that has trained hundreds of Ethiopian high school students in advanced algorithms
- Co-founded the David Harold Blackwell Summer Research Institute to increase African-American doctoral recipients in mathematics
- Served as Chair of the Computer Science Division at UC Berkeley after faculty roles at Harvard and Berkeley
- Joined Anthropic in 2026, bringing theoretical computer science depth to a frontier AI lab; honored with the ACM Eugene L. Lawler Award and PECASE
Videos & Interviews
Dr. Jelani Nelson in Conversation with Y Combinator CEO Garry Tan at the UCB Newton Series
A UC Berkeley Newton Series conversation between theoretical computer scientist Jelani Nelson and Y Combinator CEO Garry Tan, spanning algorithms research, math education, AddisCoder, and the startup world's collision with AI.
View Details
Theory of Computation with Jelani Nelson - #473
TWIML AI Podcast episode with Jelani Nelson on streaming and sketching algorithms, random projections and dimensionality reduction, what the theorist's toolbox offers machine learning practitioners, and his nonprofit AddisCoder.
View Details
