I build AI systems and the teams behind them.
I lead AI at H2 Corporation, where we are building computer vision for construction takeoff. Before that I ran applied AI at Wizard AI and spent four years as an applied scientist at Amazon. I hold a Ph.D. in computer vision from RIT.
What I am responsible for
Strategy
Setting the AI roadmap and explaining it to customers, boards and investors in terms they can plan around.
Team
Hiring applied scientists and engineers, setting the bar, and building a team that ships.
Architecture
Choosing models, orchestration and infrastructure, and making the latency and cost trade-offs that come with them.
Evaluation
Putting evaluation, staged rollouts and rollback in place early, so the team can move quickly without losing trust.
Where I have worked
Athlete Space
An AI coach for endurance athletes, which I founded in 2025 and built myself. It has analyzed 45,000 training activities for athletes in 22 countries.
More about Athlete Space
How I work
Measure before you ship
I put evaluation and staged rollouts in place before a team scales. It is the reason we can ship quickly later.
The hard part comes after the model
Most of the work in applied AI is orchestration, cost, latency and evaluation. I hire people who are comfortable with both the research and the engineering.
Explain it plainly
Boards and customers need to know what the system does, where it can fail and when it will be ready. Giving them that is part of the job.
Start from the domain
Good applied AI rests on what the experts in the field already know. I build that knowledge into the system and its tests.