BTech Computer Science and Engineering student with a focus on Machine Learning and Full Stack Development. Developed advanced ML platforms and collaborative SaaS applications, integrating technologies like Python, React.js, and AWS. Seeking a Machine Learning Engineering role to apply my skills in real-world AI applications.
Developed a Remaining Useful Life (RUL) prediction system on NASA's CMAPSS turbofan dataset (100+ engines, 20K+ sensor readings) using rolling-window feature engineering and GroupKFold cross-validation to prevent data leakage across engine lifecycles. Benchmarked Linear Regression, Random Forest, and XGBoost/LightGBM regressors, achieving an RMSE of ~15–18 cycles on held-out engines. Delivered predictions through a React fleet-health dashboard backed by a Node/Express API and MongoDB.
Built a multi-class network intrusion detection system on the NSL-KDD dataset (125K+ traffic records) to classify connections into Normal, DoS, Probe, R2L, and U2R categories. Engineered 40+ flow-based features and handled severe class imbalance via weighted training, comparing Decision Tree, Random Forest, XGBoost, and LightGBM to achieve ~97% macro-F1 on rare attack classes. Deployed the model behind a Node/Express API with a React-based SOC-style dashboard for real-time traffic scoring.