Modern enterprises increasingly depend on interconnected information systems, cloud services, software applications, third party platforms, and digital business processes. This dependence creates exposure to cyber incidents, abnormal behavior, service failures, and operational disruption. Existing cybersecurity approaches often focus on threat identification without translating technical risk into business impact or actionable response priorities. This study proposes an Explainable AI Driven Decision Intelligence framework for cyber resilient enterprise operations under uncertainty. The framework integrates cybersecurity telemetry, enterprise context, dual stream predictive modeling, service criticality, dependency analysis, Business Impact Score assessment, explainable reasoning, resource constrained prioritization, and resilience response. The predictive layer processes temporal technical signals and contextual severity attributes to estimate malicious activity and disruption probabilities. TreeSHAP and rule-based reasoning provide interpretable factors, affected services, estimated consequences, and recommended actions. The framework is designed for comparative evaluation against XGBoost, SVM, standard BiLSTM, and conventional threshold alerting using predictive, operational, explainability, resource allocation, and uncertainty measures. The study provides an integrated approach for connecting cyber threat prediction with transparent operational decisions and enterprise resilience.