Machine learning explainability frameworks often provide static, predefined explanations that fail to adapt to diverse user needs. This paper introduces a novel approach that enables users to personalize explanations through custom feature groupings, explanation goals, and context-specific preferences. Our framework allows users to logically organize model features into meaningful business-driven categories while defining the quality metrics that matter most to them. We introduce evaluation metrics including stability and diversity indices that align with user-defined objectives, bridging the gap between technical explainability and practical decision-making. Experiments with public datasets demonstrate that personalized explanations significantly improve decision-making confidence and user satisfaction compared to static approaches. The proposed framework transforms explanation from a static output to a dynamic, personalized process that adapts to specific user needs and contexts.