Kay Ousterhout
Assistant Teaching Professor at The Apache Software Foundation
Berkeley, California, United States
Summary
👤
Senior🎓
Top SchoolKay Ousterhout is an Assistant Teaching Professor at UC Berkeley with 14 years of experience building and teaching foundational computer science while designing high-performance distributed systems. She earned a PhD from Berkeley where her research and papers (SOSP, NSDI) advanced understanding of performance in large-scale analytics frameworks and low-latency scheduling. As a longtime Apache Spark committer and PMC member, she maintained core scheduler code and fixed correctness and performance issues in one of the most widely used big-data engines. In industry she led engineering at Lightstep, driving roadmaps for ~30 engineers and contributing performance-critical Go and frontend code while operating large production clusters. Kay blends rigorous academic research with hands-on production engineering, bringing uncommon depth in scheduling and observability across both systems and teaching. An interesting throughline: she repeatedly focuses on making performance behavior visible and actionable—whether in research, open source, or classroom instruction.
14 years of coding experience
13 years of employment as a software developer
PhD, Computer Science, PhD, Computer Science at University of California, BerkeleyBSE, Computer Science, BSE, Computer Science at Princeton University
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (17)
debugging10
scheduled-tasks10
debug10
spark10
java10
scheduler10
scala10
javas10
scheduling10
sched10
task-scheduler10
performance-optimization10
job-scheduling10
multithreading9
uid8
Programming languages (10)
Github contributions (5)
apache/spark
Dec 2014 - Jul 2015
Apache Spark - A unified analytics engine for large-scale data processing
Role in this project:
Back-end Developer
Back-end DeveloperContributions:3 commits, 45 PRs, 1189 comments in 7 months
Contributions summary:Kay made multiple commits focused on improving the Apache Spark scheduler, specifically addressing performance and correctness issues. The contributions involved modifying the TaskSchedulerImpl to add features like the local actor for task launching. Furthermore, the user corrected problems with task failure handling, job cancellations, and the UI by making corrections to how metrics like shuffle read time and the time to serialize the result are calculated and displayed. The main focus appears to be to enhance the reliability and accuracy of the Spark scheduler.
analyticspythondata-processingsqlapache
kayousterhout/hadoop-common
Jun 2014 - Feb 2015
Contributions:14 commits, 2 pushes, 1 branch in 8 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.