Kay Ousterhout, PhD, is interested in the performance of large-scale distributed systems. She is a co-founder of Quilt, a startup that seeks to make it easy to deploy and run software on cloud infrastructure, and a committer for the Apache Spark project.
With the support of a Hertz Foundation Fellowship, Kay completed her PhD from UC Berkeley in 2017. Kay's PhD research focused on enabling users to reason about performance bottlenecks in large-scale data analytics frameworks. She developed blocked time analysis, a methodology for quantifying performance bottlenecks in parallelized systems, and used blocked time analysis to illustrate that network and disk I/O are not as important to performance as previously believed. Given the challenges to reasoning about performance in current architectures, she explored a new architecture built specifically to provide performance clarity: the ability to understand where bottlenecks lie and the performance implications of various system changes. She demonstrated that structuring jobs as single-resource units of work called monotasks makes it simple to reason about performance without sacrificing fast runtimes.
Kay graduated from Princeton University in 2011 with a BSE in computer science. At Princeton, Kay was advised by Jennifer Rexford and Michael J. Freedman.