The Data Systems Research group at Centrum Wiskunde & Informatica (CWI) is behind several research breakthroughs ranging from compressed columnar storage to vectorized query execution that has enabled the development of modern DBMS ranging from Snowflake to Elastic. In this presentation, Peter Boncz discussed the history of the group and reviews their current research foci, including recent work on DuckDB (an OSS embeddable OLAP database initially designed to facilitate data science at scale). He also presents new research on learned data formats (to help teams manage inefficiently modeled and stored data); virtual property graphs (which identify and query graphs in relational data); and decentralized data architecture that enables privacy-preserving, distributed analytics and ML.