In an effort to enhance the performance of analytic queries in CockroachDB, a high-throughput OLTP database, the team developed a vectorized execution engine inspired by the MonetDB/X100 paper. This new engine, which improves query execution speed up to 4x, shifts from a row-oriented to a column-oriented data processing approach, enabling more efficient use of modern CPU architectures. The project involved creating a prototype to test the performance gains, employing code generation to manage the complexity of supporting multiple data types without sacrificing maintainability, and implementing batch processing to reduce the overhead of frequent function calls. The vectorized execution engine, which leverages modern CPU capabilities and reduces latency in the industry-standard TPC-H benchmark, was integrated into CockroachDB 19.2, delivering significant performance improvements for scan, join, and aggregation queries, although full implementation for all query types is ongoing.