I am interested in massive-scale distributed systems Such a system can be defined as a distributed system explicitly designed to operate in environments characterized as having many machines, usually between thousands and billions, participating simultaneously, and by having diverse networking characteristics in terms of latency, bandwidth, reliability, and generally spanning many different administrative domains. A very typical example of a wide-area distributed system is the World-Wide Web, but massive scale systems can include non-internet systems, too, such as sensor networks.

I am particularly interested in problems which relate to various forms of agreement properties, including replica consistency, consensus, and predicate detection and monitoring. Such problems are fundamentally difficult to solve in dynamic and large scale systems, due to both basic impossibility results for the strongest forms of agreement and weakest failure and synchrony models, and to scalability and performance issues, such as the sensitivity of many approaches to variability in message latency.

I received my Ph.D. from the University of Michigan where my advisor was Farnam Jahanian. Even though that was during the last millennium, I still remember what it is like to be a graduate student. I maintain a Student Resources page with links to career advice and technical resources which may be of interest to both undergraduates and graduate students.

Prospective students interested in enrolling in the Computer Science Department should read my Prosepctive Students page.