Jonathan Ullman was appointed as a Junior Fellow in 2014 as a postdoctoral research scientist in the department of computer science at Columbia University. He resigned his Fellowship in 2015 due to receiving a faculty position. He completed his Ph.D. at Harvard University in 2013 and stayed there for a year as a postdoctoral fellow in the Center for Research on Computation and Society. While at Harvard, he had the good fortune to be advised by Prof. Salil Vadhan. Prior to coming to Harvard, he graduated from Princeton 2008 and spent a year trading stock options on the floor of the American Stock Exchange.
He works on the foundations of privacy-preserving data analysis ‹ how data analysts can (and cannot) safely perform accurate statistical analyses of data containing sensitive information about individuals ‹ and its connections with fields like cryptography, machine learning, and mechanism design. His work has shown how to obtain nearly optimal lower bounds for privacy-preserving data analysis using a cryptographic primitive called a fingerprinting code, which was originally designed for watermarking digital content. More recently, he has been working on using techniques from privacy to understand the problem of false discovery in modern interactive data analysis.