Balanced Scorecard: Setting KPIs for Various Departments in an On-Demand Logistics Company

Ahamove’s rapid nationwide expansion created operational challenges, including inconsistent KPI definitions, siloed departmental reporting, and a heavy reliance on spreadsheets for performance tracking. These inefficiencies delayed decision-making and hindered cost and service optimization.

This project delivered a centralized, automated, and AI-assisted performance management system designed for managers and executives. The core components include a department-specific KPI dictionary with 40+ validated metrics, real-time Power BI dashboards with automated data refresh, and an AI Dashboard Interpreter available on Android and iOS. The interpreter enables users to upload dashboard screenshots or spreadsheets and instantly receive KPI summaries, achievement forecasts, and actionable insights via an interactive chatbot.

An experimental XGBoost classification model was developed to predict whether a department’s KPI would be met at month-end using partial-month data trained on one year of historical metrics. While this model achieved R² > 0.9 for key KPIs, it was replaced in production by OpenAI Assistant, which performs the same forecasting with greater scalability and flexibility.



The system is supported by comprehensive documentation and a governance plan for independent management. Future enhancements include expanding KPI coverage, integrating advanced anomaly detection, enabling batch analysis, and connecting directly to operational databases for live KPI tracking.