Delivery Performance Dynamics Analysis

Analyzed a real-world delivery dataset to understand what drives performance and customer satisfaction. Using OLS regression and ANOVA, identified statistically significant correlations between personnel metrics, delivery times, and satisfaction scores — separating signal from noise in a multi-variable environment.

Built machine learning models in Python, including Random Forests, to predict delivery time and rank the factors with the most impact on overall efficiency. The final model achieved an F1 score above 0.85, demonstrating strong predictive accuracy across the test set.

The project highlighted how operational decisions — staffing, routing, and workload distribution — compound in their effect on end-to-end delivery performance.