Tanmoy Saha

Engineered a predictive model Predict Blood Donations, improving donation prediction accuracy by 85% using Python and real-world datasets. Optimized the data preprocessing pipeline, reducing data cleaning time by 30% through automated feature engineering and efficient missing value handling. Visualized donor behavior patterns using Matplotlib and Seaborn, increasing actionable insight extraction efficiency by 40%. Designed and delivered interactive dashboards for stakeholders, enhancing data-driven decision-making effectiveness by 25%. Experimented with AutoML TPOT and Logistic Regression models, achieving an ROC-AUC of 0.7891 and improving overall predictive reliability across test samples.