I work on optimizing storage systems through analyzing and exploiting data access patterns, and learning the relationship among the stored data. This falls under the broader scope of the Cognitive Storage initiative which seeks to automate data life cycle management, resource provisioning, tiering, caching, prefetching, etc. in an intelligent way.
Prior to joining IBM, I worked on distributed algorithms for network throughput optimization, database viewsize estimation, and characterizing access patterns for high energy physics data as a postdoc at INRIA. I studied mathematics and statistics for my undergraduate degree at Seoul National University, and later received my PhD in operations research from UC Berkeley.