Shelton Bumhe

I trained a machine learning model to predict if a quarterback would throw a touchdown in an NFL game. It pulls in stats like recent performance and trends, and reached about 88% accuracy.

I also created a web app using Streamlit so users can upload stats and get predictions, with SHAP visualizations to explain how it works.

I built a Java program that helps companies manage their rentals, purchases, and leases. It connects to a MySQL database and generates reports. I handled the full backend logic and wrote JUnit tests to verify it all worked smoothly.

I wrote a Python tool that solves knight's tour puzzles using recursion. It visualizes every move as the knight tries to cover the entire board — without repeating any square.