Developed a complete end-to-end machine learning pipeline covering data ingestion, preprocessing, model training, evaluation, and inference, following a modular and production-oriented project structure.

Implemented feature engineering and preprocessing pipelines (categorical encoding, numerical scaling) and trained a regression model to predict student performance, with trained model and preprocessor persisted as serialized artifacts.

Utilized CatBoost-based modeling with experiment artifacts and logs stored locally, enabling reproducible training and consistent inference across runs.

Deployed the trained pipeline through a Flask-based web application, allowing users to input student attributes and obtain real-time performance predictions via an interactive interface.