Fast and adaptive logistics are essential for meeting evolving customer expectations and needs. In response to recent shifts in consumer behavior, which have altered the perceived value of products, companies, and services, organizations are increasingly turning to intelligent algorithms to swiftly adapt to various scenarios. Cross-docking, a strategy known for enhancing flexibility and reducing delivery times, is central to this approach. This paper introduces an advanced AI-driven framework utilizing an evolutionary algorithm to tackle the multi-dock truck sequencing challenge in cross-docking centers, aimed at minimizing makespan. Our method includes a machine learning-based parameter tuning strategy, enabling the algorithm to automatically adjust to different instances without human intervention. We evaluated our framework by comparing results obtained with and without machine learning tuning against state-of-the-art methods. The findings indicate that our AI-enhanced approach not only delivers superior feasible solutions but also achieves these results with reduced computational times, positioning it as a highly effective tool for logistics optimization.