Reddit Hackathon
New York, NY
Developed a simple, addictive recall game for Reddit’s Devvit Mini Hackathon, 'AI Builders x Reddit Games'
For the mini hackathon, I used Bolt to vibe code a simple food court tray recall game—designed as a quick break from passive scrolling to test memory and spark a little nostalgia.
Landing Page Background (JavaScript)
Washington, DC
Perlin noise equation to produce reactive visual.
Uses a simple Perlin noise equation, angle mapping (0–2π), and mouse-distance physics to control particle motion. Each dot starts at a fixed location and is wiggled by smooth noise, pulled dynamically toward the mouse, and redrawn with ghost trails.
"Unprecedented Transnational Growth and Rampant Capitalism within the Religious Economy"
Columbia, South Carolina
Revealed aspects of the religious economy.
Abstract: This essay analyzes how the bait-and-switch organization centered around a fear-inducing-evil-galactic-overlord, Xenu, his spacecraft, and exploding volcanoes manages to overcome dozens of external imperatives including litigations for extortion, coercion, tax evasion, and falsification of records...
Automated file converter (PowerShell)
Washington, DC
Developed a script to convert hundreds of PowerPoint files into PDFs.
Saved hours for my team and avoided an Adobe subscription by runnning an automation over all .ppt files in a folder.
Data cleansing algorithm (Stata)
Columbia, South Carolina
Extracted thousands of messy files from Refinitiv then used Stata to clean and organize for Python ingestion.
Processed thousands of disparate datasets from archived Refinitiv data 1962-2023, creating a structured dataset with all investments by country and year, tagged with data such as round number, investment stage, and deal vallue.
Global health investment file loop (Excel Macros)
Columbia, South Carolina
Cleaned and structured massive datasets for global health investment trends.
Wrote an Excel Macro to loop through thousands of files and compile data into a single large dataset. This was for the purpose of comparing Refinitiv data with enterprise investment data (including project team, project industry, type of product, etc). My work exposed which factors contributed most to a successful investment across the investor small teams.
Bad actor algorithm (Python)
West Point, NY
Worked on a small team to create an anomaly detection algorithm that runs continuously based on user inputs.
Improved the usability of the Army Big Data Platform using a backend based in the theory of k-means cardinality to detect abnormalities in daily data points collected in large quantities.
Disease detection algorithm (Python)
West Point, New York
Used machine learning to identify biological patterns in disease.
Modeled biological markers using a multilabel transformation and support vector classifier. I achieved .02334 logistic loss. More info: https://www.kaggle.com/competitions/lish-moa