About

ReferralExchange

I'm Head of Data Engineering at ReferralExchange. I lead the data platform and design systems for entity deduplication, lead and referral enhancement, large-scale transaction retrieval, and contact scoring. I also lead agentic coding practices across engineering.

Work and projects

MusicAI (late 2018–2021). I built and launched it as CEO and Head of AI. We predicted emerging artists from raw audio alone.

We ingested about 50,000 new songs across the web daily, processing them on one computer I built with NVIDIA 2080 Ti GPUs and lots of RAM. AWS cost under $1,500 a month. It was my first computer build; I learned Linux, GPU computing, TensorFlow, PyTorch, Athena, and S3 along the way.

We worked with Universal, Warner, and Sony Music, helping discover and break Dominic Fike, Tones and I, and Remi Wolf, and helping artists and labels choose album tracks for release.

CensusIQ.

We built for 14 months, raised just shy of $1 million, and later declined VC funding.

L7V.

Founding engineer on Alpha, a startup validation tool. Our team went from idea to working product in 54 days. What I learned.

Get Match Fit.

I built the app, website, ML models, and social automations, turning football performances into fan workouts. Over one million organic social views in 30 days; about 15,000 weekly active website users as of October 2026.

The Music Atlas.

An interactive D3 map of Spotify genres and tracks, built at USF.

Breakout Mentors AI Academy.

I founded and led it (2019–2024), wrote the curriculum, and managed 7–15 mentors.

quiltomics.

Founding data scientist and technical staff lead. I led a team of four building genomics databases, pipelines, and visualizations.

I graduated cum laude in data science from the University of San Francisco.

Research

Research I've coauthored:

MuSLCAT.

Music modeling from raw audio. arXiv preprint, 2021.

Song Hit Prediction.

Predicting Billboard hits using Spotify data. arXiv preprint, 2019.

HeartBioPortal.

Cardiovascular genomics data. Circulation: Genomic and Precision Medicine, 2019.

Music genre classification.

Deep learning from raw audio. Research poster, USF Creative Activity and Research Day, 2019.

Poway Wrestling

I'm president and CEO of the Poway Wrestling Foundation and a 2016 Poway Wrestling alumnus. With our board and staff, I led:

Finances.

From $7,000 in debt to over $100,000 annual surplus and over $250,000 in the bank in three years. Annual revenue: $75,000 to nearly $500,000 in three and a half years.

Athlete growth.

In three years, elementary/middle school grew from 25 to around 150; high school from around 80 to over 150. Our senior Olympic division has around 20 athletes.

Team.

Multiple world and Olympic medalists, our largest coaching and team staffs ever, and nutrition and strength training with The Treigning Lab.

Life abroad

I lived in Mexico for a year during COVID, then spent another year living in seven countries across the Middle East and Asia, including Vietnam and Japan.

I moved to Tokyo independently, opened my own company, and secured my Business Manager visa through it, without outside sponsorship.

Interests

AI and agents.

Ontologies, better recall, fewer hallucinations, and agents working together.

Building and distribution.

I used to avoid selling. Now marketing excites me too.

Teaching and sport.

Helping people learn and improve through Breakout Mentors and Poway.

Life in Tokyo.

My wife, books, and ramen with friends.