B.Tech · NIT Calicut · Class of 2026
AI, ML & Data Enthusiast | Python · Java · Scikit-Learn · XGBoost · LightGBM | Building production-ready pipelines
End-to-end ML pipelines — from raw data and feature engineering through to REST-API-ready inference schemas.
I write modular, reproducible code: serialisable pipelines, clean input/output contracts, independently testable components.
| Project | Stack | Key Result |
|---|---|---|
| Customer Churn Prediction | Logistic Regression · SMOTE · GridSearchCV · Platt Scaling | Calibrated binary classifier; Elastic Net + SAGA |
| Cirrhosis Outcome Prediction | XGBoost · Stacking · Scikit-Learn | Multi-class log loss; 5-fold CV |
| Retail Demand Forecasting | LightGBM · ARIMA · Croston · STL | MASE 0.82–0.89; adaptive model selection |
