A Systematic Review of AI-Driven Business Intelligence Architectures for Data-Informed Strategic Decision-Making Methods (2019–2026)

Abstract

This study conducted a quantitative systematic review of AI-driven business intelligence (BI) architectures to evaluate their effectiveness in enabling data-informed strategic decision-making between 2019 and 2026. A total of 142 peer-reviewed studies were systematically identified and analyzed, revealing a 130% increase in publications from 2019 (n = 12) to 2024 (n = 28), indicating rapid growth in this research domain. The dataset showed that 58.4% of studies employed quantitative methods, 29.6% used mixed approaches, and 41.5% relied on large-scale datasets exceeding 10,000 records. Primary findings demonstrated that AI integration improved system efficiency by an average of 31.5%, reduced processing time, and enhanced predictive accuracy by 18.7%, with 74.2% of studies reporting significant accuracy gains. Decision-making outcomes improved substantially, as 71.1% of studies reported enhanced decision quality and a 27.3% reduction in decision latency due to real-time processing capabilities. Regression analysis indicated that AI integration explained 42.5% of the variance in organizational performance, with strong beta coefficients exceeding 0.60 for key variables. Sub-group analysis showed that finance (26.1%) and healthcare (22.5%) sectors achieved the highest predictive accuracy improvements above 22%, while supply chain (30.1%) and retail (28.4%) sectors demonstrated greater operational efficiency gains. Effect sizes ranged from 0.65 to 0.75, confirming moderate to strong impacts across performance indicators. Overall, the findings provided robust quantitative evidence that AI-driven BI architectures significantly enhanced efficiency, analytical accuracy, and strategic decision-making effectiveness across industries and regions.