In the fall of 2019 I built a platform for The Syllabus to send endlessly customized "syllabi" to their subscribers every week. It launched in late September and is building steam. (Flask, Pydantic, AWS S3 + SES, Vanilla JS)

In late 2018 at the Recurse Center I made a prototype of a human-machine transcription interface I like to call Transcribely. It's pretty amazing, in that it saves upwards of 50% of labor compared to a human transcribing from scratch. (React.js + Flask) Contact me for a demo! It'll soon be in production at Averbach Transcription.

In 2015 a couple friends and I made Tappt at a hackathon. It's a web app that allows users to search for keywords inside videos. We've since fleshed it out into a commercial service. (Python, Flask, JS)

In the fall of 2017 I designed, implemented, and deployed a work assigner for Averbach Transcription. This is run as a background process, assigning transcription work automatically based on transcribers' current workloads. It saves Averbach Transcription 30+ hours of labor every month in assigning and re-assigning transcription projects. (Python)

In the summer of 2016 I designed, implemented, and deployed automatic invoice generation for Averbach Transcription. This integrates with Freshbooks web API v2.1, generating invoices from completed transcription projects. It saves Averbach Transcription 20+ hours of labor every month compared to the old manual invoicing process. (Python)