A Senior Designer's Claude System for Prototypes That Match the Real Product
Learn cross-functional UX design, product management & AI skills at: uxcel.com
Your AI prototypes look generic because something is missing before the prompt, and it's not the model's fault.
That's exactly what we explored in this episode with Carmen Rincon, senior product designer and co-founder of Yummy Design Studio & Labs, a studio that helps scale-ups and growth-stage teams when design becomes the constraint. Across 10 years she's designed for AT&T and Nissan at VOX Global, worked in the regulated world of energy at CleanChoice Energy, led product design at Learn.xyz, and was the founding designer at an AI startup in San Francisco. She also runs Yummy Labs, a live two-week accelerator where early-career designers solve real startup challenges.
Carmen's clients aren't coming to her because they fear AI. In her words: "AI is the bottleneck, not the designers."
What you'll learn:
- Why startups now bring designers in after AI, not before, and what that fragmentation pattern looks like
- The "first-week hire" mental model that changes the quality of everything AI gives you
- How to build design tokens Claude can actually use (it reads semantics, not pixels)
- Why a design-system MD file beats calling the Figma MCP every time: tokens, crashes, speed
- Carmen's full client pipeline: Figma → Claude project → Storybook → an npm package devs update with one line in the terminal
- Where to start tomorrow without overcomplicating it
Timestamps:
00:38 What this episode covers
01:56 The skill Carmen is focused on now (knowing when not to use AI)
03:58 Career path: PR firm, energy industry, founding designer
08:30 Why she started building AI design workflows
14:39 Why AI prototypes look generic
19:01 Live demo: design system for a Claude-built travel app
28:26 Claude projects and the design-system MD file
32:27 Prototyping micro-interactions without Figma's prototype feature
36:45 Where to start tomorrow
Where to find Carmen:
- LinkedIn: https://www.linkedin.com/in/carmenerincon/
- Yummy Design Studio & Labs: https://www.yummy-labs.com/
- Yummy Labs on LinkedIn: https://www.linkedin.com/company/yummy-labs/
Resources and tools mentioned:
- Claude and Claude Code (claude.com) — AI assistant and coding agent; ran the whole setup, from design-system MD files to the Storybook build
- Figma (figma.com) — where the tokens, variables, and component descriptions live, including Carmen's honest verdict on the Figma MCP
- Storybook (storybook.js.org) — open-source UI component workshop; how the design system gets handed to developers
- Gluestack — the client's UI library; Claude Code translated the whole system into it
- npm — the package pipeline: Carmen pushes a Figma change, devs update with one line in the terminal
Brought to you by:
Uxcel is a learning platform used by 500,000+ product professionals and 200+ organizations to build cross-functional skills across UX design, product management, and AI. Learn through short, gamified courses that fit into your busy schedule.
Your AI prototypes look generic because something is missing before the prompt, and it's not the model's fault.
That's exactly what we explored in this episode with Carmen Rincon, senior product designer and co-founder of Yummy Design Studio & Labs, a studio that helps scale-ups and growth-stage teams when design becomes the constraint. Across 10 years she's designed for AT&T and Nissan at VOX Global, worked in the regulated world of energy at CleanChoice Energy, led product design at Learn.xyz, and was the founding designer at an AI startup in San Francisco. She also runs Yummy Labs, a live two-week accelerator where early-career designers solve real startup challenges.
Carmen's clients aren't coming to her because they fear AI. In her words: "AI is the bottleneck, not the designers."
What you'll learn:
- Why startups now bring designers in after AI, not before, and what that fragmentation pattern looks like
- The "first-week hire" mental model that changes the quality of everything AI gives you
- How to build design tokens Claude can actually use (it reads semantics, not pixels)
- Why a design-system MD file beats calling the Figma MCP every time: tokens, crashes, speed
- Carmen's full client pipeline: Figma → Claude project → Storybook → an npm package devs update with one line in the terminal
- Where to start tomorrow without overcomplicating it
Timestamps:
00:38 What this episode covers
01:56 The skill Carmen is focused on now (knowing when not to use AI)
03:58 Career path: PR firm, energy industry, founding designer
08:30 Why she started building AI design workflows
14:39 Why AI prototypes look generic
19:01 Live demo: design system for a Claude-built travel app
28:26 Claude projects and the design-system MD file
32:27 Prototyping micro-interactions without Figma's prototype feature
36:45 Where to start tomorrow
Where to find Carmen:
- LinkedIn: https://www.linkedin.com/in/carmenerincon/
- Yummy Design Studio & Labs: https://www.yummy-labs.com/
- Yummy Labs on LinkedIn: https://www.linkedin.com/company/yummy-labs/
Resources and tools mentioned:
- Claude and Claude Code (claude.com) — AI assistant and coding agent; ran the whole setup, from design-system MD files to the Storybook build
- Figma (figma.com) — where the tokens, variables, and component descriptions live, including Carmen's honest verdict on the Figma MCP
- Storybook (storybook.js.org) — open-source UI component workshop; how the design system gets handed to developers
- Gluestack — the client's UI library; Claude Code translated the whole system into it
- npm — the package pipeline: Carmen pushes a Figma change, devs update with one line in the terminal
Brought to you by:
Uxcel is a learning platform used by 500,000+ product professionals and 200+ organizations to build cross-functional skills across UX design, product management, and AI. Learn through short, gamified courses that fit into your busy schedule.