About Me
"We do these things not because they are easy, but because we thought they were going to be easy."
Hey, I'm Justin! I'm a passionate and dedicated professional who started my AI learning journey in May 2024. Since then, I've been committed to mastering this exciting field through extensive self-driven coursework and projects. I've completed programs from notable institutions such as Stanford University and OpenAI, utilizing online platforms like Coursera and Kaggle. I hold two Bachelor of Science degrees from the University of California San Diego: Electrical/Computer Engineering and Mathematics/Economics.
Equipped with hands-on experience in deep learning, programming, and data science, I am adept at handling complex problems and implementing innovative solutions. My career goals center around making a significant impact in AI, especially in financial markets, gaming, and large language models. I am constantly exploring new ways to harness this accelerating technology to create positive value and transformative insights.
Latest Projects
Kalshi Trading Bot AI
I designed an automated trading bot for Kalshi prediction markets that implements the multi-variable Kelly Criterion to optimize bet sizing across correlated markets. The bot continuously monitors market orderbooks, calculates optimal portfolio allocations considering correlations between events, and dynamically manages limit orders. It features risk management through fractional Kelly sizing, market depth analysis for effective pricing, and automated order execution via the Kalshi API.
View Project
Particle Life AI
I developed an artificial life simulation based on Particle Life where autonomous particles interact based on simple rules of attraction and repulsion. The particles resemble cellular automata and exhibit behaviors such as self-organization and pattern formation. The simulation demonstrates how complexity can emerge from simplicity and is written in JavaScript, allowing users to run it directly in the browser.
Play Project
MTG Card Generator AI
In this project, I constructed a Magic: The Gathering card generator. Initially programmed in Python, it was later adapted to JavaScript for integration into my personal website. The script first randomizes attributes such as card type, rarity, mana cost, and color. Then it uses these parameters to dynamically generate a new MTG card.
Play Project
Tetris AI
I built an AI that plays the classic game Tetris using a neural network trained through evolutionary algorithms. The AI considers all possible moves then uses a neural network to evaluate the resulting structure using these input features: lines cleared, maximum height, minimum height, number of holes, bumpiness. Techniques like tournament selection and elitism were used to improve performance.
View Project
Snake AI
In my Snake AI project, I created an AI that plays the game Snake using a neural network trained with an evolutionary algorithm. The AI learns to navigate, avoid collisions, and maximize its score by eating food. Inputs include: nearby obstacles, food location, snake direction, snake length.
View Project
Flappy Bird AI
For my first game AI project, I made an AI to play Flappy Bird using NEAT (NeuroEvolution of Augmenting Topologies). I trained the AI to control when the bird jumps. Inputs include: bird's height, distance to pipe, height of gap. The project features both a human-playable version and an AI-controlled version.
View Project