I've worked on all sorts of projects in my life. These are the ones for which the following evaluates to true, for my choice of epsilon.
(time_spent > 2 months) || (is_interesting && on_github && 100 - percent_finished < epsilon)
Capstone Project with PNC *
Python
PNC was very scrupulous about its data protection!
For this reason, too much detail can't really be divulged.
However, I can say that in working with and advising industry partners, we deployed models and tools
ranging from basic TF-IDF models to Word2Vec and BERT based embeddings. And importantly, this project polished off my Master's coursework.
Vulnerability Detection *
Python, C++, C, Assembly
While at Qualcomm, I worked with the Linux Security Team on a multi-prong project.
Using recurrent neural networks, I helped build a model which could detect vulnerabilities in assembly code.
To improve the accuracy of the network, I independently developed a modular symbolic execution tool using angr
Lastly, I used a Docker container on a Kubernetes cluster to set up the tool for corporate deployment.
Database from Scratch *
C++
To complete a major in Computer Science at Carnegie Mellon, one must choose from a list of daunting finale classes,
advanced in their content but also practical in their design. For Database Systems I learned modern theory
on database efficiency and implemented it throughout the semester in C++.
Fast Matrix Multiplication (paper)
Python
For a class final project, I delved into the latest research on theoretical matrix multiplication runtime bounds. You can find that paper here.
Advanced Matrix Algorithms
Python
Over the course of a semester, my graduating class for the MS-DAS program implemented advanced computational matrix algorithms from scratch. You can find those here.
Neural Network from Scratch
Python
As part of one of my graduate classes, and to ensure I really understood how feed forward networks operate, I wrote it myself.
You can find that code here.
Personal Python Package
Python
After rewriting the same helper functions time and time again, eventually I decided to just package it up nicely. You can access the package here, and learn more about what it includes as well.
Atlas Resonate *
Swift
During an internship with Audition Technology,
I worked with the Mobile Development Team to develop the Atlas Resonate app. I rapidly acquired the relevant skills in Swift and XCode,
and subsequently became the primary iOS developer. The app interacts with Apple's Healthkit and employs advanced data analysis to analyze hearing health data.
You can download the current version of the app here.
Gravity Simulation
Python
A lifetime ago (well, high school), I wrote a simulator to see how gravity would differ in more/less than three dimensions for the science fair.
The original code is lost to history, but I decided to
redo the project, this time with some visualization! You can view the code
and a basic explanation of the theory here.
Snake Game
Python
In an interview with Lex Fridman, machine learning and artificial intelligence researcher Charles Isbell from Georgia Tech cautions programming novices not to dip their feet in the coding deep end.
He says endeavoring to accomplish something great programmatically before learning the basics will almost surely lead to frustration, and eventually giving up.
But after years of perfecting the skill, Isbell sat down to write Pacman in a day, with the hopes that it would be trivial after all this time. Indeed it was.
This is my Pac-Man.
Meme My Playlist
Python
Memes.
Memes.
Memes.
There has to be a way to automate this. Oh wait, REST APIs!
P.S. You need a (paid) developer account to make this into an actual Apple Music playlist.
If someone finds this project interesting enough, I will accept a sponsorship.
More to come!
I'll continue to keep this site updated as I work on bigger and better projects in the future.
* Either prioprietary or includes academically sensitive content, so some details are/code is ommitted.