Computer Engineering and Computer Science at Northeastern University

I started programming at eight with small Arduino projects. Since then, that curiosity has grown into work across computer vision, machine learning, embedded systems, and application development.

A privacy-first macOS window manager that uses short-lived context signals to surface the right apps at the right time. It is built in Swift and designed to process context in memory without retaining clipboard contents or behavioral history.

An empirical study of how three detector families behave under unseen generators, adversarial paraphrasing, and domain shift. The project compares feature-based, neural, and zero-shot approaches and documents where confident predictions fail to generalize.

A real-time computer-vision pipeline that recognizes intentional blinks and turns them into configurable macros. The system combines eye-region features, sequence modeling, live inference, and on-device processing.

A deep reinforcement-learning agent trained with Soft Actor-Critic to reach orbit in Kerbal Space Program. The repository includes the training work and written project report.

• NURobotics: Member since 2026 as a programmer on the marine robotics team.
• FRC LBL Team 7403: Member from 2019 to 2024 and programming lead in 2024, guiding the software work behind the team's competition robot.
• TOM Medellin: Maker and systems lead from 2022 to 2024. Helped develop a smart cane that detects objects up to six meters away, a WhatsApp chatbot for local farmers, and an internal tool that tracked service hours for more than 100 makers.
• Club Campestre Foundation: Led a student development team in building an internal system for scholarship distribution and benefactor management.
• PACO: Worked part-time as a software developer in 2025.
• Harvard Pre-College Program: Built image-classification models that achieved more than 94% accuracy on CIFAR-10 and MNIST in 2023.
• Climate education: Created an interactive Python and Arduino game in 2025 to help children explore how wildfires affect climate change.

For machine learning and vision, I usually work with Python, PyTorch, scikit-learn, and OpenCV. For applications and systems, I use Swift, Java, Dart/Flutter, C, and C++. I also enjoy rapid hardware prototyping with Arduino and building lightweight services with FastAPI.