Large-scale technology programs operate under persistent uncertainty arising from schedule interdependencies, cost pressure, technical complexity, changing requirements, governance constraints, and the coordination of multiple projects and stakeholders. Conventional risk-management processes remain indispensable, but periodic reviews and qualitative assessments may not fully exploit the time-stamped operational data generated throughout program execution. This study develops a longitudinal predictive risk-management framework designed to identify adverse technology-program conditions before formal risk escalation and translate predictive information into managerial decision support. The proposed empirical setting consists of major U.S. federal IT investments observed through 2024. The unit of analysis is the investment-reporting period. The primary outcome is a subsequent adverse CIO risk condition within a prespecified prediction horizon, with rating deterioration and severe cost or schedule variance evaluated as secondary outcomes. Predictor variables include current and lagged cost and schedule variance, temporal trends, volatility, spending dynamics, project concentration, prior risk trajectory, and scale indicators. Logistic regression, decision trees, random forests, and gradient boosting are evaluated using chronological validation and an independent out-of-time test set. Performance criteria include recall, precision, F1-score, ROC-AUC, PR-AUC, Brier score, calibration, and warning lead time. A real-data feasibility pilot of 50 quality-screened completed project records shows meaningful variation in cost and schedule performance while revealing the importance of explicit data-quality controls. Final predictive performance is intentionally not fabricated and must be populated after execution of the full historical panel. The framework contributes a replicable architecture connecting operational program signals, predictive risk intelligence, explainable model outputs, and human-governed management interventions.