I build the big systems organizations trust — and I make them feel human.
Thirty years of building software, from Fortune 100 AI to tools for small nonprofits. The throughline: technology that respects the people using it — especially the people who usually get the worst tools.
Built for mission-driven organizations
Software your team can actually trust
The kind of operational software that runs an organization — built to be loved by the people who use it every day, not just tolerated.
Co-founded with Tom · CRM + donor + volunteer management
GoodCloud
The all-in-one platform that ran small nonprofits.
A full CRM with donor management, volunteer coordination, and donations — everything a small nonprofit needs to run, in one place. Tom and I built it in four months, open-sourced a stack of libraries from it, and made a lot of small organizations very happy.
The same world as the work you're considering — built with the same partner who introduced us.
Accessibility & human-centered design
Built for the people who get the worst tools
Designing for real human beings — including people navigating disability, illness, or brains that work differently. This isn't a specialty I added. It's been the work since 2003.
UCSD · Medical research
Joint Count Project
Accessible health tracking for patients with advanced arthritis.
An accessible joint-pain reporter built for patients living with advanced arthritis — genuinely usable for them, and producing clean, doctor-friendly output on the other end. Peer-reviewed research came out of the work.
Accessible software for people with disabilities — more than twenty years ago.
The Good Years Project
Bank of Health
Treat your health like wealth.
A personal health app that turns decades of longevity research into a simple ledger: see exactly how many healthy hours each choice — a workout, a drink, a call to a friend — adds or subtracts from your life.
Dense science, turned into something anyone can use.
Four years building high-stakes AI
AI that augments people, instead of replacing them
I've built AI at the highest levels, and the goal has always been the same: make the humans smarter and faster, never replace them. If you're wondering how to get your team using AI well, this is the part to read.
Technical Lead · Fortune 100
Valence
AI leadership coaching, trusted in the Fortune 100.
Led engineering, architecture, and UX to turn a rough proof-of-concept into a secure, scalable, production-ready AI coaching platform used inside Fortune 100 companies — blending generative AI with traditional software so it was reliable enough to trust.
Serious organizations trusted this in production.
Head of Engineering · Government
Ox Intel
Beautiful AI decision-making for high-stakes calls.
An AI decision-making tool built to augment human judgment on hard, high-stakes questions — used by federal government and cities, and engineered to run even on air-gapped, hard-to-reach servers.
Government-grade trust, without the government-grade ugliness.
Open source · 2013
Will
A friendly, teachable AI assistant — a decade before ChatGPT.
One of the first cross-platform chatbot assistants, open-sourced and loved by developer teams. Easy to teach, easy to talk to, and genuinely helpful — the friendly-AI-for-normal-people instinct, years before it was a trend.
Approachable AI has been the instinct all along.
Live · AI research tool
Front of the Napkin
A real answer to "should I pursue this idea?"
Trade a few dollars for genuine research on an idea — competitors, search demand, pricing, market gaps — turned into a clear report. Agentic AI doing real work, wrapped in a simple, human interface.
Modern AI applied usefully, not as a gimmick.
And for thirty years, serious organizations have trusted me to build and to lead.
Wieden+Kennedy
Creative technical lead — Coke, Nike, Laika, and a Super Bowl spot.
Honda
An e-learning framework that beat out 10 firms and ran unchanged for six years.
BuddyUp
Co-founder & CTO. Real-time chat at scale across iOS, Android, and web.
Open source
12+ libraries — Will, Salad, Polytester, Inkshop, and the django-* family.
Who I am
I am not a personal brand. I'm a person.
I've built a lot of things. Some succeeded, plenty failed — and I keep a public list of all of them, because the world would be better if more of us were honest about what doesn't work.
If your organization is doing real good in the world and trying to figure out where technology — and AI — actually fit, I'd love to talk.