Kyan Cox

Hi 👋

I'm Kyan Cox.

Aspiring

Resume

About

Hello! I'm Kyan — a Computer Science student at UW-Madison with a passion for software development.

I've loved programming ever since writing my first line of code in high school. Whether it's creating software to solve problems, exploring the possibilities of artificial intelligence and machine learning, or diving into the world of cryptocurrency and blockchain technology, I'm always excited about building and learning.

Outside of coding, I love listening to music. Explore my recent music taste below, powered by the Spotify Web API!

Education

University of Wisconsin - Madison

Majors: Computer Science, Statistics

Expected: May 2026

GPA: 4.0

Currently playing

Top tracks this month

Spotify data provided by Spotify Web API.

Experience

September 2024 - Present

UW-Madison College of Engineering

AI/ML Undergraduate Researcher

- Collaborated with a team of 10 peers to develop machine learning models, including gradient-boosted trees, identifying key factors impacting donation campaigns for The River Food Pantry, the largest food pantry in Wisconsin.

- Applied predictive analytics and SHAP analysis to evaluate model accuracy and interpretability, offering detailed insights into feature importance, enhancing donor engagement strategies, and improving resource allocation for the non-profit organization.

- Occasionally volunteering at The River Food Pantry, helping with manual labor while also meeting with directors to discuss application, allocation, and usage of donation data.

Python

Machine Learning

Scikit-Learn

October 2024 - November 2024

UW-Madison Interdisciplinary Professional Programs

Student Facilitator

- Facilitated hands-on activities and provided guidance to industry professionals, including engineers and managers, in understanding and applying Machine Learning and Artificial Intelligence concepts to their careers.

Machine Learning

Artificial Intelligence

Python

Technical Instruction

May 2024 - August 2024

Outlier.ai

AI Training (Contract)

- Evaluated AI-generated code using RLHF techniques to enhance LLM accuracy.

- Developed and implemented test cases for Python and JavaScript, ensuring high-quality AI performance.

RLHF

Python

Java

TypeScript

JavaScript

May 2023 - June 2023

Rye Chamber of Commerce

Intern - Software

- Simplified a multi-day process of extracting and cleaning data from spreadsheets through Python and the Pandas library, aiding in the identification of over 200 potential customers, ultimately maximizing funds for events.

- Developed a mail merge process in Python to create outreach envelopes for prospective customers.

Python

Pandas

Git

Excel

Projects

DARS Visualizer

DARS Visualizer allows UW-Madison students to view their DARS reports in a more organized manner, including charts and graphs to make reports easier to understand.

Description:

- Created a Next.js frontend, integrating Chart.js to dynamically visualize degree completion data from UW-Madison DARS reports, enhancing student understanding of academic progress.

- Developed a FastAPI-based web service for parsing and analyzing reports, utilizing PyMuPDF for PDF text extraction, regex for pattern matching, and Pydantic for data validation.

- Containerized the application with Docker and deployed it on AWS EC2, demonstrating proficiency in cloud deployment and API development.

Next.js

React

Chart.js

FastAPI

AWS (EC2)

Docker

Tailwind CSS

TypeScript

Python

More info →

Visual Snow Log

A mobile app designed for those with Visual Snow Syndrome (VSS), a rare condition that affects ~3% of the world including myself. Visual Snow Log allows users to log and track their VSS symptoms over time, and export them to their email to show their doctor.

Description:

- Implemented user authentication and data storage using Supabase and PostgreSQL.

- Created a backend API with FastAPI and Python to handle data processing and export functionality, utilizing Pandas for data manipulation and the smtplib library for sending emails via SMTP.

- Deployed the backend on an AWS EC2 instance, containerizing the application using Docker for consistent development and deployment environments.

React Native

Expo

FastAPI

Pandas

AWS (EC2)

Docker

PostgreSQL

Supabase

TypeScript

Python

SQL

More info →

LLM Comparison App

A web app that allows users to compare responses from leading Large Language Models (LLMs) like ChatGPT, Gemini, and Claude. Input a prompt, generate responses, and vote for your favorite to see how each model performs.

Description:

- Developed a full-stack web application using Next.js, React, and TypeScript to compare responses from OpenAI’s ChatGPT, Google’s Gemini, and Anthropic’s Claude, utilizing their respective APIs.

- Integrated MongoDB for data storage and management, and utilized Node.js for server-side operations, including API endpoints and middleware to enforce query limits for premium models, ensuring efficient and fair usage.

- Implemented a voting system and data visualizations using Chart.js and Tailwind CSS.

React

Next.js

Tailwind CSS

MongoDB

Node.js

TypeScript

Chart.js

More info →

Cryptocurrency Portfolio Allocator

A web app where you can view all of your cryptocurrency assets from Coinbase, Gemini, and Ledger Live in one place. The app utilizes your API keys from your Coinbase and Gemini accounts, as well as a CSV file of your assets from Ledger Live.

Description:

- Created a RESTful API using Flask, featuring 14 endpoints to manage customer information within an SQLite database, that seamlessly integrates three external APIs for dynamic data retrieval, and employs OOP principles to ensure scalability and a maintainable code architecture.

- Utilized Pandas to both create an XLSX file for customers that exports their assets from multiple services into one centralized portfolio and parse through Ledger Live’s Operation History CSV.

Python

Flask

SQLite

JavaScript

HTML

CSS

Pandas

RESTful API

More info →

Skills

Programming Languages