I build AI products for decisions that matter. I take ambiguous, high-leverage problems from strategy to working software, with governance, explainability, and human judgment built into the product rather than added after.
My work spans AI-native developer platforms, advertising, financial services, insurance, and operational software — the kinds of systems where a small mistake can carry real financial, compliance, or customer consequences.
I prototype directly in code and with AI development agents rather than working only from slides. I review around working product. I move fast, default to writing, and own the tradeoffs across speed, correctness, and trust.
I lead cross-functional teams across engineering, design, ML, GTM, and executive stakeholders in regulated enterprise environments, doing the player-coach work of driving the highest-leverage initiative while raising the craft bar for the team around me.
Trust is a product behavior, not a disclaimer. People should be able to see what the system knows, what it inferred, and who is responsible for the decision.
Footing is the product I’m taking furthest: a working system for commercial contractors that keeps the estimate, review, proposal, sent record, and outcome connected. Teams can bring in the work they already use, see what changed, keep decisions tied to the right version, and move forward without hiding uncertainty behind one tidy answer. I’m building it directly with contractor feedback, from product strategy and governance through the working software.
Lumen is a working product prototype exploring how brands, creative teams, and media owners can create ads, set no-code requirements, route automatic and human review, compare fictional performance scenarios, and preserve the reasoning behind every decision. It grew out of a question I kept coming back to from retail media: how do you make creative move faster without making accountability harder to see?
Interactive prototype using fictional advertisers, sample forecasts, and simulated approvals.