The group has been defining important analytic extensions to SQL since Oracle 8i. We cooperate with other database vendors to standardize our SQL extensions so that they are adopted quickly. We are responsible for developing innovative algorithms for optimizing and executing the proposed extensions in the RDBMS kernel. Many of our algorithms have been patented and our work has been published in major conferences. We are responsible for providing scalability of sql operators for parallel queries and for memory usage. A most recent area of interest is vectorized and columnar processing. We extended the SQL language to identify data that is closest match to query vector. We generate optimal query plans for retrieve data fast. We take advantage of new on-disk and in-memory storage formats and hardware caching characteristics to improve join, aggregation, sort, window and hierarchical query performance. We build and enhance new access and filtering structures, such as, bitmap index, bloom
