I'm a software engineer based in Chicago. I have worked as a machine learning or software engineer at several early-to-mid-stage startups: XaiPient, Zero Hash, and Synthflow. I also worked for a year at the Research Computing Center at the University of Chicago, where I developed software for researchers, maintained web applications used by researchers, and advised researchers on software issues related to their grant proposals.
Two of my favorite professional accomplishments are that I placed in the top ten in a Kaggle competition and that I created a popular open source library with dozens of contributors.
Prior to working as a machine learning and software engineer, I obtained my BS in Mathematics from Tulane University and my MS in Statistics from the University of Chicago. My favorite subject in math was set theory, and I had the opportunity to study the axiom of determinacy at the UCLA summer school in mathematical logic.
Recently, I have become more interested in functional programming languages like Haskell and Purescript due to their higher level of type safety and their use of mathematical language and concepts.
You can read about some of the open source projects I created on the "Projects" page.
Here are a few of the open source projects that I've created over the years:
A JavaScript library that allows you to prompt users for microphone permissions and run callbacks on segments of audio with user speech in a few lines of code. Under the hood, it runs a voice activity detection model in the browser. You can see a demo of it here.
A computational life experiment that runs in the browser, created in PureScript. It is an implementation of the paper Computational Life: How Well-formed, Self-replicating Programs Emerge from Simple Interaction by Blaise Agüera y Arcas et al, which I discovered in an episode of the Mindscape podcast. It's not entirely finished, but the basic functionality is there.
A general framework for dictation and speech-based input. The user writes a Python script that defines how patterns of speech map to Python functions to run, and these functions can leverage the Python ecosystem to simulate keystrokes or issue commands to applications. It is based on the academic work surrounding kleenexlang. In my opinion, it is one of my most creative projects, although its usability is limited by the quality of open source speech-to-text models that was available at the time that I created it.
A website showing the most mentioned cities on r/digitalnomad. The visualization is deployed at spikynomadball.ricky0123.com.