• Improved the accuracy of the dropout prediction model by 20% using SQL for data preprocessing and Power BI forvisualization, leading to better informed decision-making.• Encoded 6 categorical variables, enhancing model performance and providing more meaningful insights for strategic planning.• Elevated model performance by 10% and reduced false positives by 25%, leading to more accurate predictions and cost savings.• Provided actionable insights to academic advisors, enhancing 90% retention strategies.