Development of an AI-Driven Predictive Decision Support Model for Multi-Domain Business Analytics and Infrastructure Optimization

Abstract

This study addresses the persistent organizational problem that many firms possess growing volumes of business and infrastructure data but still lack an integrated AI-driven predictive decision support model capable of converting those data into high-quality decisions and optimization outcomes across multiple domains. Its purpose was to develop and empirically examine a unified model linking predictive analytics, business analytics capability, data integration, organizational readiness, decision quality, and infrastructure optimization within a quantitative, cross-sectional, case-based research design. Using survey data from 210 respondents drawn from cloud-enabled and enterprise-oriented operational contexts across operations, IT/data systems, finance/planning, infrastructure or asset management, and strategy or administration, the study measured six core variables: AI-Driven Predictive Analytics, Business Analytics Capability, Data Integration Capability, AI Decision Readiness, Decision-Making Quality, and Infrastructure Optimization. The analysis plan combined descriptive statistics, reliability testing, correlation analysis, and multiple regression modeling. Findings showed strong construct reliability with Cronbach’s alpha values ranging from 0.83 to 0.89, while all major variables recorded high mean scores, including AI-Driven Predictive Analytics at 4.18, Business Analytics Capability at 4.09, AI Decision Readiness at 4.05, Decision-Making Quality at 4.21, and Infrastructure Optimization at 4.12. Correlation results were positive and significant, with the strongest association found between Decision-Making Quality and Infrastructure Optimization at r = .71, p < .001. Regression results indicated that the model explained 56% of the variance in Decision-Making Quality and 62% of the variance in Infrastructure Optimization. AI-Driven Predictive Analytics significantly improved Decision-Making Quality (beta = .34, p < .001), while Decision-Making Quality was the strongest predictor of Infrastructure Optimization (beta = .38, p < .001). Overall, the study concludes that AI-driven predictive decision support becomes most effective when analytical capability, integrated data, and organizational readiness are aligned, producing measurable gains in decision quality and infrastructure performance. The implications are that organizations should treat AI decision support as an integrated capability system rather than as a standalone technical tool.