Manoj AA

Led end-to-end retail analytics pipeline using Python (Pandas) by cleaning and integrating multi-source datasets, and engineered key metrics like Incremental Revenue % and Sales Uplift to evaluate promotion effectiveness across campaigns and stores, identifying high-impact products and enabling data-driven business decisions. Architected an end-to-end ML model to predict customer price segments using behavioral and demographic data, leveraging XGBoost and Random Forest with advanced feature engineering. Shipped a Streamlit app with real-time inference to enable actionable pricing and customer segmentation decisions.

Architected cohort-based regression models by segmenting customers on age and risk profiles, improving premium prediction reliability by 30% and reducing manual validation time by ~2 hours per dataset. Enhanced data preprocessing pipeline using VIF, advanced encoding, and outlier handling, improving data integrity by 38% and reducing cleaning effort by ~1 hour per cycle. Optimized XGBoost models with GridSearchCV, improving prediction accuracy from 73% to 98% while minimizing extreme errors to 2%, ensuring stable and scalable premium estimation.