The traditional database query optimizer, a cornerstone of relational database management systems for decades, operates on a fundamentally limited premise: it makes decisions based on static statistics, predefined heuristics, and a constrained exploration of potential execution plans. This approach, while effective for simpler workloads, buckles...