Brendan Lee - Projects

TRAI (Trading AI)

Developed a trading platform that evolved from a web scraper for stock-related content to a full-stack solution integrating sentiment analysis and market predictions. Currently implementing collaborative trading features to enable group-based investment strategies.

Live Demo

Full-Stack Flask React Node.js MongoDB Web Scraping Sentiment Analysis Natural Language Processing TensorFlow NLP Financial Data

IMC Prosperity 3

Achieved top 0.14% globally (17th out of 12,000 teams) in IMC's algorithmic trading competition. Engineered a trading bot implementing market making, mean reversion, and statistical arbitrage strategies across multiple asset classes.

Can't be giving away our alpha that easily :).

Python Statistical Modeling Data Analysis Trading Strategies Backtesting Game Theory

Connect Four Bot

Built a game-playing AI using the minimax algorithm with alpha-beta pruning and Q-learning to achieve competitive Connect Four gameplay.

ConnectFourAI

Python Minimax Algorithm Game Theory Numpy TensorFlow PyTorch Matplotlib

Stock Web Scraper

Fetches real‑time market data, pulls the latest Yahoo Finance headlines, scores each story’s sentiment, and lets an LLM rank the news by its likely impact on the stock. Currently training TextBlob library on financial news sentiment data for better results.

StockWebScraper

Python BeautifulSoup4 Requests yFinance Pandas

Physics-Based Soccer Simulation

Developed a comprehensive physics engine from scratch in C to power a realistic soccer simulation game. Implemented collision detection, force dynamics, player AI with team coordination, and realistic ball physics with accurate momentum transfer during player interactions.

SoccerPhysics

C SDL Physics Simulation Collision Detection Vector Mathematics Game AI

Bitcoin Price Predictor

Used regression and XgBoost to predict future prices of bitcoin from closing prices. Working to build system to detect arbitrage opportunities in the crypto market and competing in DRW's cryptocurrency Kaggle compeition.

BitcoinPredictor

Python Pandas NumPy Scikit-learn Matplotlib Seaborn XgBoost

Memory Visualizer

Working on a program in HTML, CSS, and JavaScript to show C heap/stack memory allocation. Inspired by Caltech course CS 3 (Software Design).

MemoryVisualizer

HTML CSS JavaScript

Machine Learning Projects

Comprehensive portfolio of machine learning implementations spanning deep learning, supervised learning, and predictive analytics.

ML

Python Pandas Numpy TensorFlow PyTorch Scikit-learn Matplotlib Seaborn Data Analysis Data Visualization Statistical Modeling

Competitive Coding

Competitive coding problems from USACO, CodeForces, CSES, and Advent of Code.

CompetitiveCoding

C++ Data Structures and Algorithms Problem-Solving

Personal Website

Built a responsive personal website using HTML, CSS, and JavaScript to showcase my projects, skills, and experience.

PersonalWebsite

HTML CSS JavaScript

Python Statistical Tool

Working on a statistical tool package in Python. Inspired by two Caltech courses: ACM 11 (Introduction to Computational Science and Engineering) and ACM 157 (Statistical Inference).

StatisticalTool

Python Numpy Statistics