Andrew Quinn

I'm a sixth year Ph.d candidate working with Jason Flinn and Peter Chen at the University of Michigan. I'm supported by a Microsoft Research Fellowship and a National Science Foundation Fellowship.

I'm broadly interested in improving software reliability. Specifically, I'm investigating data-intensive software reliability: how can we apply data-intensive techniques to prevent, detect, and resolve software errors? My thesis proposes a fundamentally more powerful model for observing program executions to perform tasks such as security forensics, debugging and data provenance. This work treats a program execution as a massive data object consisting of all the program states (i.e., memory and register values) reached by the execution and uses cluster-scale parallelization and declarative languages to accelerate and simplify observations over these objects. In the future, I'm interested in using similar techniques to provide run-time safety properties with minimal developer effort.

I am currently mentoring a number of students in Baris Kasikci's lab. We're working on a number of projects in the software reliability space, including improving the reliability of persistent memory applications, understanding how heterogeneous deployments impact the reliability of our systems, and building better tools for failure reproduction.

I'm an avid reader, political junkie, and casual sports fan. If I'm not working, then I'm probably wrestling with my dogs, Huckleberry and Harley; cooking an elaborate meal with my wife, Tully; baking something sweet from Bravetart; or enjoing a long run around Ann Arbor.