From Black Box to Executable Logic: Explainable Reinforcement Learning through Prolog Expert Systems | AIChainDay