Abstract
Diabetes mellitus (DM) is a heterogeneous metabolic disorder characterized by chronic hyperglycemia resulting from defects in insulin secretion, insulin action, or both. Globally, the prevalence and economic burden of DM continue to rise, necessitating new approaches for prevention, diagnosis, clinical management, and health‐system planning. Data intelligence (DI)—the integrated use of data engineering, analytics, machine learning (ML), and artificial intelligence (AI) within robust socio-technical systems—has emerged as a transformative enabler across the diabetes continuum of care. We synthesize evidence on key application domains, highlight clinical and operational outcomes reported to date, and analyze barriers related to data quality, algorithmic bias, privacy, interoperability, and real-world implementation. We propose a pragmatic evaluation framework and a research roadmap focused on explainability, causal inference, hybrid mechanistic–ML models, and equitable deployment at scale.
