Brady McAtee

Software Engineer · Full-Stack Web, Machine Learning, 3D Graphics

Download Resume (PDF)


About

I started programming at 12 making games in Visual Basic, and I fell in love.

Right now I'm studying CS at Portland State with a physics minor and working as a software engineer intern at MSEI/Biotronik, where I build simulation models for medical device hardware.

I like working on interesting engineering problems and building useful technology, whether that involves computer vision, 3D graphics, machine learning, or full-stack web development.

Projects

recipeBox — Recipe management web application

Built a full-stack web app for restaurants to manage recipes and ingredients with full CRUD support. Designed RESTful APIs and normalized PostgreSQL schemas to keep recipe and ingredient data consistent. Implemented responsive React UI with intuitive forms for adding, editing, and searching recipes.

GitHub · Live Demo

Stack: React, Node.js, Express, PostgreSQL, AWS EC2

Tennis Shot Tracker — Ball tracking and serve velocity measurement

Trained a ball tracking CNN model based on the TrackNet architecture using PyTorch. Used computer vision techniques and homography estimation to map 2D video coordinates to real-world court dimensions. Measured tennis serve velocity from iPhone video recordings by combining frame analysis and geometric calibration.

GitHub

Stack: Python, PyTorch, Computer Vision, OpenCV

Interactive Photo Refocusing App — Depth estimation and DOF simulation

Built full-stack app using Python, PyTorch, and Flask for DSLR-style depth-of-field simulation from single images. Implemented Multi-Scale CNN (87.9M parameters) achieving 79% accuracy for monocular depth estimation. Created responsive web interface with real-time click-to-focus interaction and configurable aperture effect.

GitHub Live Demo

Stack: Python, PyTorch, Flask, Docker

Plant Disease Classifier — ML disease detection

Created PyTorch/ResNet model achieving 99% accuracy on 50,000+ plant images. Used data augmentation to improve model performance across plant species. Deployed model on Hugging Face Spaces for real-time disease diagnosis.

GitHub · Live Demo

Stack: Python, PyTorch, ResNet, Hugging Face

Ray Tracing Engine — 3D graphics renderer

A ray tracing engine built from scratch in C++ that renders realistic 3D scenes. Implements reflection, refraction, shadows, and anti-aliasing.

GitHub

Stack: C++, 3D Graphics, Computer Graphics

Skills

Languages C++, C, C#, Python, JavaScript, SQL
Web React, Node.js, Express, Flask, PostgreSQL
ML/CV PyTorch, OpenCV, NumPy
Tools Git, Docker, Linux, Vim, .NET, AWS

Contact

Email: bradymcatee12@gmail.com

GitHub: github.com/bradymcatee

LinkedIn: linkedin.com/in/brady-mcatee