A data science capstone project from my final semester at Purdue University. We built and compared multiple classification models (Logistic Regression, SVM, KNN, Random Forest) to predict Purdue's performance in the 2024 NCAA Tournament, evaluating them against both business objectives and traditional data science metrics.
A serverless AI-powered running coach that generates a personalized two‑month rolling training plan, adapts weekly based on performance and feedback, and answers training questions in real-time. Built on modern web + AI patterns with cost-aware rate management.
A cloud-native ETL pipeline for ingesting, transforming, and analyzing financial market data using AWS, GCP, Docker, and Apache Airflow. This project is currently under active development and serves as a hands-on demonstration of modern data engineering and DevOps practices.