Leading a stellar team and continuing to build Sodatone post-acquisition. Applying data and deep learning to scouting, marketing, and new initiatives across WMG.

Acquired by Warner Music Group. Sodatone helps record labels discover, evaluate, and track high-potential artists, powered by deep learning and structured data across 250M songs, 10M artists, and 1B short-form videos.Artists discovered and signed through Sodatone have resulted in 40B+ streams, and several Top 10 chart positions worldwide.

Cardiogram uses deep learning to predict cardiovascular conditions such as diabetes, atrial fibrillation, and high blood pressure from Apple Watch data. We're Andreessen Horowitz-backed and a team primarily of ex-Google machine learning engineers with advisors from OpenAI and UCSF.

Worked on LinkedIn's Talent Solutions team, building a new machine learning-driven recruitment product.

Architected and shipped desktop control software for atomic force microscope (MEMS AFM) from scratch. Brought-up and validated signal conditioning PCBs, and built real-time data collection, processing, and visualization libraries.

Designed and implemented motion planning and computational geometry algorithms, converting PCB design files into 4D toolpaths. Wrote 50% of circuit board printer codebase, active on first $500,000 of Kickstarter units.

Wrote and recorded 4 EPs with my high school punk band 🤘https://phasmaband.bandcamp.com/album/detour