My research interests include cloud computing, storage and distributed systems and operating systems. My graduate research has focused on two closely related directions: 1) making MapReduce instances more efficient, when running on both bare metal clusters and virtualized clusters, through prudent resource management and dynamic performance tuning; and 2) improving storage efficiency in virtualized data centers by designing content-aware de-duplication based caching protocols and a similarity-aware virtual machine management mechanism. I am enthusiastic about applying system principles to solve practical, real-world system problems and to build deployable systems.