Today, companies need to plan for skills need, risks to employees and the gaps in skills and capabilities, and achieve productivity and long-term growth while doing so.Strategic workforce planning is becoming more important for companies to consider and know what skills they might need, what risks may be facing their employees and what skills might be missing. In this study, an AI-based Workforce Planning framework using Extreme Gradient Boosting (XGBoost) with Python and Jupyter Notebook is proposed. The framework combines employee, skill, performance, engagement, training, workload, productivity and workforce-demand indicators to categorize employee profiles for strategic workforce interventions. A reproducible analytical workflow is integrated and includes data pre-processing, train-test partitioning, model training and performance evaluation. An illustrative synthetic data set of 6,000 workforce records is used to demonstrate the proposed method without misrepresenting the information that is available for organizations. The model XGBoost was able to reach an accuracy of 81.8% and a ROC-AUC of 0.898, demonstrating discrimination between stable and intervention required profiles. Skill gaps, role criticality, workforce demand, engagement and workload were found to be key influencing factors through feature-importance analysis. The framework enables evidence-based decision making for the sustainable growth of organisations related to reskilling, retention, recruitment and deployment.