Hi! I am a software engineer with a focus on machine learning and the music industry.
I started my career in 2013 of internships for NASA, the US National Security Agency, and the US Department of Energy’s Oak Ridge National Laboratory (ORNL). At ORNL, I researched ways that quantum computing could theoretically accelerate Bitcoin mining.
In 2015, I joined Uber as one of the youngest engineers they had ever hired, where I worked on Uber's public-facing developer API. Shortly before I left, I led the development of a prototype of what later became UberHEALTH—Uber’s non-emergency medical transportation product.
I left Uber in 2017 to start a now-failed startup. I went back to school, did some work in a computer security research lab, and then started another company in late 2019.
Here is where you can find my long-form writing on the Internet.
- Click Track, a newsletter analyzing the future of the music business.
- Dark Shift, a pop culture publication.
- Operational Security for Activists, a free online book that explains how activists can stay secure when using technology. I wrote this for my friends in the aftermath of the 2020 George Floyd protests.
- machinelearning.wtf, an online encyclopedia of machine learning terms. Currently accepting contributions via GitHub.
These are some of the major open source projects that I have started.
- ScalarStop, a framework for keeping track of machine learning experiments.
- Provose, a new way to manage your Amazon Web Services infrastructure. You describe the containers, databases, and filesystems that you want to deploy, and Provose automatically calculates the necessary security and networking configuration. Built with HashiCorp Terraform.
- TFCA, a HashiCorp Terraform module that makes it easy to create a local self-signed TLS Certificate Authority.
- Ori, a Python library that provides high-level concurrency tools. For example, the Ori PoolChain makes it easy to distribute workloads across many threadpools and process pools.
- Cypunct, a Python library that makes it easy to quickly split Unicode strings based on entire Unicode character classes. This is useful for rapidly tokenizing text for natural language processing.