Zero to one design for AI products.

The hard part of a first release is not drawing screens. It is deciding what the product actually is, what earns a place in version one, and what gets cut so the thing can ship. That decision work is most of the value here.

Six weeks, end to end, from a broad idea to a scoped production quality design your engineers can build from. This is the same approach I have used designing a banking platform, a commercial real estate deal platform, and a defense situational awareness interface, applied to AI native products.

Weeks 1 to 2. Discovery and definition. Workshops to turn the idea into something buildable. Who it is for, the jobs it does, the core journeys, and a prioritized feature set with explicit reasoning about what ships first and what waits.

Weeks 3 to 5. Product design. Every core screen designed to production quality: information architecture, navigation, interaction patterns, responsive behavior, and the AI specific states most first releases forget. Generation, streaming, confidence, correction, and failure.

Week 6. Handoff. Specifications, component documentation, and working sessions with your engineers so the build starts without a design bottleneck.

A buildable design, and the reasoning behind it. The second part matters more than teams expect. When your team makes the next hundred decisions without me, the documented reasoning is what keeps those decisions consistent with the first hundred we made together.

Founders and product teams taking an AI product from concept to a first real release, with engineering ready or nearly ready to build.

Teams who want screens drawn to an existing spec. If the thinking is already done and you need production output, the retainer is a better fit and cheaper.