- 🎓 Studying Computer Science + Business at Northeastern University
- 💻 Web & app developer focused on clean, user-centered products
- 🤖 Exploring machine learning with TensorFlow & Keras
- 🎮 Building interactive experiences in Unity 3D
- 🎯 My goals: master coding → build AI → make robots → start my own company
- 🤝 Clubs: Northeastern Entrepreneur Club · AINU Club · NU Robotics Club · Hackathons (Chatathon)
-F2C811?style=for-the-badge&logo=powerbi&logoColor=black){kind=icon}
📋 Click to view full skill breakdown
| Category | Skills |
|---|---|
| Languages | Java · Python · C++ · HTML · JavaScript · Kotlin |
| Mobile Development | iOS (Swift) |
| Machine Learning | TensorFlow · Keras |
| Hardware & Embedded | Embedded Systems · Raspberry Pi 5 |
| Cloud | Microsoft Azure |
| Data & Analytics | Power BI (DAX) · Microsoft Excel |
| Design & Media | Figma · Photoshop · Adobe Premiere Pro |
| Game Dev | Unity 3D |
| Dev Tools | Git / GitHub · Visual Studio Code |
| Productivity | Microsoft 365 · Microsoft Word |
🤖 PiCar-X Lego Turret Robot — joystick-controlled servo turret on a Raspberry Pi 5 robot car
Built a joystick-controlled servo turret and Lego missile launcher on a SunFounder PiCar-X from scratch, with no existing tutorial to follow.
- Fixed SPI communication failures by wiring the joystick straight to the Robot HAT V4's built-in 12-bit ADC pins, which removed an external chip, and tuned PWM on pin P3 for precise turret pan
- Wrote one Python control loop (
combined_app.py) that combines mobile app control, live WiFi camera streaming and real-time servo control- Published wiring diagrams, code and a demo video
Tech: Python Raspberry Pi 5 Robot HAT V4 Embedded Systems | 🔗 Repo ·
🧠 Image Classification with ML & Deep Learning — Python Developer, University of Arkansas at Little Rock (Summer 2025)
Built image classification models during a summer role at UA Little Rock, including a parking-space detector for real-time camera use and a CNN that reached 95% validation accuracy.
- Parking-space detector: a scikit-learn pipeline that labels spaces as empty or occupied, using HOG features on images resized to 15×15 and an SVM tuned with GridSearchCV
- Deployment: stratified 80/20 train–test splits, with the trained model saved as a pickle file so it can classify new camera images on the fly
- Deep learning: a CNN built in Keras/TensorFlow and trained on a 25,000-image dataset with data augmentation, reaching 95% validation accuracy and evaluated with confusion matrices
Tech: Python scikit-learn TensorFlow Keras NumPy Pandas Matplotlib Seaborn | 🔗 Repo
🔹 PROJECT NAME THREE — short one-line description
What it does, the problem it solves, and your role.
Tech: Unity 3D C# | 🔗 Repo
I'm always open to collaborating on interesting projects, internships, and co-op opportunities. Feel free to reach out!
