I am primarily responsible for the tiering management and garbage collection features of our distributed data store. More generally, I have worked on the data store as a whole including tuning the data path, migration, and QoS code. I am a first level resource for finding and fixing memory leaks and other performance related bugs.

In addition to the C++ work I do on the data store, I also have had a hand in our Python test infrastructure, and other Python automation tools.

Major focus on investigating potential storage savings in indexing systems for non-hierarchical file systems. Additional work in studying the efficacy of performing a computation versus storage trade-off in the cloud; methods of modeling/simulating this trade-off, and possible use cases.