Degrees
- Master of Science in Artificial Intelligence [In Progress]
- Bachelor of Science and Arts in Mathematics + Certificate in Elements of Computing
- Associate of Science in Mathematics
- Associate of Arts in French
Positions
- Associate Security Software Engineer at Indeed
- Full-Stack Web Developer at Komak Solutions
What I'm looking for
Role in DevOps, machine learning / AI, data science, or some other SWE role. Preferably remote or in Austin, Texas.
Software engineer, mathematician, outdoorsman, piano player, oud player, former martial artist, trilinguist, and neophyte woodworker.
I come from a strong mathematics background (huge math nerd) and am pursuing a career in software engineering, with a long-term goal
of moving toward AI. When I'm not on the job, I thoroughly enjoy playing music and engaging in outdoor activities like hiking, skiing, and
fishing. In life I strive to combine my passions; I'm currently working on a project where I apply calculus and numerical methods to model
the oud, using software engineering to create a more efficient way to build the instrument. I'm testing my method by actually
building it — hence "neophyte woodworker".
During my sophomore year as an undergrad, I spent a couple of months fervently teaching myself JavaScript, HTML, CSS, and MongoDB.
Just seven months after I started programming, I was employed by Komak Solutions, a tech consulting firm in Texas, where I worked
as a freelance full-stack developer for six months. It was my first taste of the tech industry, and I loved it.
This thirst for knowledge continued into my junior year when I signed up for a data science boot camp with General Assembly. This
10-week intensive course equipped me with strong data science skills in Python and statistics. My final project was an analysis of
7000 foreclosed properties, where I built a data model to predict the profitability of property flips. The instructors even
suggested that I could work as an instructor for General Assembly.
Later, I applied my data science knowledge to an independent project where I analyzed ECG data from over 10,000 patients, building
machine learning models to diagnose heart conditions. These experiences solidified my passion for data science and machine learning.
Most recently, I worked as an Associate Security Software Engineer at Indeed for about 2 years, where I gained some great experience
in IAM and learned a lot of the industry's technologies. I aim to expand my skillset in various areas, including DevOps, machine
learning, and general industry expertise. Currently, I am pursuing a Master of Science in Artificial Intelligence at UT Austin, with
a long-term goal of advancing in the AI field. Ideally, I would love to apply AI expertise in the medical/biotech field to profoundly
impact people's lives.
In this independent analysis of ECG data from over 10,000 patients, I built multiple ML models that diagnose various heart conditions.
The classifiers I used were KNN, decision trees, and logistic regression. My best model was a decision tree that can diagnose a patient
with an abnormal heart rhythm, with an accuracy of 94% and a recall score of 0.93.
See the PDF here.
In December 2022, I started playing the oud. By 2024, I became interested in building one. After watching many
videos on traditional, time-consuming methods, I analyzed the process mathematically and developed a more efficient
approach. I invented a mathematical model for the oud/lute bowl profile, and devised a method that leverages calculus
and numerical techniques to rapidly generate accurate rib templates for ouds/lutes of semi-circular cross-section.
I created a website that allows users to customize the bowl shape using my model, then generates the rib template: oudrib.com
See the math paper explaining my method here.
For the final project in my Deep Learning graduate course, my group and I trained a neural network in
SuperTuxKart Ice-Hockey using DAgger imitation learning, creating an AI that competes with other agents.
Future enhancements may incorporate reinforcement learning to improve strategies. This project highlights
the transformative potential of machine learning in gaming.
See the PDF here.
Squeezer is a webapp that ranks stocks by their potential for being short-squeezed. It uses webscraping and APIs to get stock data, then uses an algorithm that I developed with the help of a finance major to assign a 'squeeze index' to each stock. My main responsibilities included the Home page, the Time Machine functionality, the ranking algorithm, and general design.
In my undergrad senior year, I took a fascinating numerical analysis class and decided to build a library of implementations of the concepts we learned. This is one example from my library: Newton's Divided Differences interpolation is a method for generating a polynomial that is guaranteed to intersect a given set of points.