I build design systems to design better products
Digital products that are cohesive, intuitive, and consistent from start to finish. My work lives as much in the product's structure, its logic and what it solves, as in the experience and interface that reach the user.
Selected work
Design system for a B2B AI platform
In production for five years. From tokens to UI and AI components, documented for humans and machines.
Controlled Generative UI
AI-generated analyses inside the design system. The AI picks from 6 artifact types following DS rules; the frontend renders real components, auditable by default.
AI-native Design System (personal)
A system whose documentation was designed from day zero to be consumed by humans AND AI — no retrofitting.
What I do
Recent notes
- Controlled Generative UI: how to make AI pick charts without breaking your DSin progress
The pattern, the schema, and the rules — applied in a B2B HR product in production.
- 5 semantic palettes for data viz — a taxonomy distilled from ColorBrewerin progress
Sequential, categorical, divergent, status, highlight. Why these five and not more.
- Why Design System documentation needs to be machine-readable in 2026in progress
Designing docs for humans AND for the AIs that will help your team consume them.
I've designed for fifteen years. I started in graphic design and branding, then moved through agencies, public institutions and private companies, doing product, brand and web. In 2020 I went fully into product design. Since then I architect products and design systems so larger teams execute with autonomy.
My specialty is the foundations that let products scale, the tokens, components, patterns and documentation. In 2026 that work runs through AI in the process, with machine-readable docs, versioned prompts and workflows that cut repetitive work by 40 to 50%.
About →How I work
Obsidian + RAG · Agent crew (Claude + MCP) · Figma · Tailwind v4 · Storybook · Fumadocs
Full ecosystem →