Flows, screens, components, motion, states. Built to ship, not to screenshot. Every decision is documented so engineering can build without guessing.
I design products the way engineers build them, in systems, not screens. Every component, every state, every motion is documented. The output is a Figma library that mirrors the codebase, plus a handoff package engineering can read on day one. The product is yours to extend, defend, and ship without me.
5-8 stakeholder interviews, 3-5 user interviews, content audit, competitive scan. Output: research synthesis and opportunity map.
Sitemap, user flows, edge cases. Tested on paper with 3-5 users. Output: signed-off architecture.
Low-fidelity wireframes, all key flows, all major states. Two rounds. Output: wireframes that survive contact with reality.
High-fidelity screens in Figma, built on the system grid. Every component is named, versioned, and documented.
Framer prototypes for the 3-5 critical flows. Motion specs for every transition. Tested with 3-5 users.
Spec document, weekly implementation reviews, 2 rounds of QA. The design survives contact with engineering.
Pick the scope. The price is the price. No hourly, no discovery fee, no surprise add-ons after kickoff.
The three to five flows that actually convert. Wireframes, high-fidelity screens, and a clickable prototype, scoped to the surfaces that matter, not the whole product.
The product designed as a system. Flows, screens, components, motion, and an engineering handoff that can be built from on day one.
Research through implementation QA. User interviews, the full product system, motion, and two rounds of engineering review so the design survives contact with code.
The questions specific to PRODUCT that come up in every strategy call. Honest answers, no pitch.
The ideas behind this track, written down in the operator log.
SaaS product design is not a moodboard exercise. It is an operating system: research, flows, trust states, conversion surfaces, and a design system that survives contact with engineering. Here is how I run it end to end.
Learn how to design AI-powered SaaS products that users actually trust, covering uncertainty communication, control design, error handling, and onboard…
AI features fail users when trust, memory, and recovery are treated as backend concerns. Here are the interface patterns that keep AI SaaS usable when the model is slow, partial, or wrong.
Learn how to design AI products that handle uncertainty gracefully-with UX patterns for confidence levels, error states, and output variability that bui…
Brand to URL.
Your next project starts
with a 15-minute diagnosis.
Available for 2 new projects this quarter.