Matheus Fonseca

Software Engineer with strong foundations in Computer Science principles due to CS50x, Codepath Tip102, and self-study. Fluent in algorithms and data structures, I strive to think outside the box, inventing new ways to solve problems. In fact, my favorite word is "why": "why does this work like it does?", "why do we do it this way instead of this other way". This mindset causes me to approach every problem with curiosity which I consider the heart of learning.

Projects

SendToMac

A cross-platform application that lets users transfer text and links from iOS to macOS, even if the Mac is offline. Built with Python, AppKit, Rumps, SwiftUI, it features real-time notifications, scheduling/expiration logic, and secure REST API communication for reliable and seamless cross-device workflows.

Python SwiftUI REST API AppKit
Github

Chess

An interactive web application that enforces all chess rules, including check, checkmate, stalemate, and legal move validation. Designed with HTML, CSS, and JavaScript, it features modular move logic, real-time piece interaction, and a responsive UI for smooth and accurate gameplay.

HTML CSS JavaScript
Site

Group Hackathon Projects

Balatest

ByteHacks, CCNY Hackathon

A Balatro-inspired educational card game that transforms study materials into interactive boss battles. Upload your notes to generate question cards with varying difficulties. Correct answers deal damage to the boss while wrong answers cost HP, with failed retries incurring additional penalties in this strategic study experience.

Python Flask Gemini API
Showcase

NeuralFeedback

SBU Hacks, Stony Brook Hackathon

An AI-powered product feedback tool that generates multiple reviewers with unique personalities and demographics based on user-specified criteria. Built with the Gemini API, it simulates diverse customer perspectives to provide targeted critiques on product ideas, highlighting strengths, weaknesses, and demographic-specific concerns.

Python Flask Gemini API Eleven Labs
Showcase

GreenMindAI

HackMIT

An eco-conscious AI chatbot that routes queries through three models (TinyLlama, Hugging Face, Claude) based on carbon emissions. The system prioritizes the most sustainable model first, falling back to more powerful options only when necessary, using token count and CodeCarbon metrics to minimize environmental impact.

Python TinyLlama Hugging Face Claude API
Showcase

Contact Me

Feel free to reach out to me directly or use the form below!

Email: [email protected]