Led end-to-end development of a full-stack editorial management platform, from initial requirements gathering with journal editors to deployment. Drove sprint planning and iterative delivery, conducted code reviews, and facilitated user testing sessions to validate workflows against real publication needs.

A cross-platform React Native/TypeScript app that resolves women's jean sizing inconsistencies across brands like H&M, American Eagle, and Aeropostale. Users input their size from one brand and get accurate conversions for others, powered by a normalized Firebase database and an ML-based fit recommendation engine. Architected with a clean service layer and Vitest unit tests covering core conversion logic.

Developed a crowdsourcing mobile application that allows users to report and track police presence at MBTA train stations in real-time. The app calculates probability scores based on user submissions and provides historical data visualization, enhancing commuter safety awareness and providing peace of mind when navigating unfamiliar transit locations.

Developed an educational platform using React and Supabase designed to reduce student dependency on AI tools through interactive comprehension assessments. Integrated Claude AI and Google Cloud Speech-to-Text APIs to provide real-time evaluation of student understanding and generate actionable feedback for educators, promoting deeper learning engagement.

A terminal-based word puzzle game written in Java where players fill a customizable grid with letters to spell words. Each round, the player receives a letter and strategically places it within the grid; completed words clear from the board, making room for new ones. Built around object-oriented design with clean separation between game state, grid logic, and I/O — and uses binary search against a word list to validate words instantly as they form.

Developed during my Object-Oriented Design course, this game showcases advanced object-oriented programming concepts and practices. It incorporates design patterns such as decorators, adapters, and strategies, while adhering to the MVC (Model-View-Controller) architecture.

Developed a machine learning model in Python to analyze the correlation between weather conditions and air quality across major US cities. Utilized APIs for data collection, performed extensive data cleaning (Pandas, NumPy), and created visualizations (Matplotlib) to enhance interpretability. Implemented and compared multiple ML models (Scikit-Learn) to optimize accuracy.

Developed a tool that compresses PNG images by removing seams of pixels with the lowest energy, improving image processing efficiency and maintaining visual integrity.