I am a research assistant and a PhD candidate at the Distributed Systems Lab (DSL) at UCSB. I am interested in distributed systems, data storage, data management, data privacy, cloud management, and data mining research.
Data Management: My work focuses on using data placement to enhance the performance of distributed data management systems on multiple scales. One project, GPlacer, focuses on the placement of geo-replicated databases (databases that are replicated in multiple geographical regions to tolerate datacenter scale outages and enhance read latency) to enhance transaction latency without violating the application requirement. Another project, HotCache, addresses the problem of load balancing in highly scalable distributed caches. This project aims to dynamically achieve load balacing between caching servers under highly scaled and skewed workloads without affecting the long tail lookup latency.
Data Privacy: I am also interested in building systems for data privacy. First, we developed TaoStore, the first tree-based ORAM store that supports asynchronous and concurrent accesses. Then, we presented Pharos (EDBT 2018 best demo award), a demo that highlights the privacy hazards of the naive usage of asymmetric locks and quorum in replicating ORAM stores. In addition, we build systems that help social network users to have more control over their privacy. Our vision is that data noise must be added from the source to protect social network users from attribute (e.g., location, gender, ethnicity, etc) inference attacks. For this, we presented LocBorg (SIGSPATIAL 2017 Second Place Best Vision Paper Award), a system that sits in the social network user's machine to help her defend her privacy.