Abstract
The growing complexity of workforce management in the merchandising industry requires an integrated and intelligent human resource system. Traditional HR processes that rely on manual encoding or separate digital tools often cause delays, data redundancy, scheduling conflicts, and limited decision-making support. To address these issues, this study developed Merchandising 1 Human Resource III, an AI-based workforce optimization system using scikit-learn. The system integrates Attendance, Scheduling, Timesheets, Leave, and Claims Management into a unified platform designed to improve efficiency, accuracy, and workforce planning. The study applied Agile Scrum methodology in developing the system using a microservices-based architecture with API integration for real-time data synchronization. Machine learning algorithms such as decision trees, clustering, and regression were implemented through scikit-learn to analyze attendance records, employee availability, and historical workforce data. The system was evaluated through functional testing, integration testing, and user assessment involving managers and employees. The system automated attendance monitoring, schedule generation, timesheets, leave processing, claims management, dashboards, and access controls. Predictive analytics were described as improving schedule accuracy and reducing conflicts, while users reportedly benefited from greater transparency. No sample size, scheduling-accuracy value, conflict count, time-saving measure, model-validation metric, or user-rating result was reported. The integrated platform may support centralized workforce administration and future scalability. Claims of improved fairness, allocation, accuracy, and efficiency require defined evaluation measures and comparative data. The specific machine-learning models selected and their performance should also be reported.
