About me
Designer who vibe codes,
builder who designs.
👋 I'm Suhee Lee, a Senior Product Designer with 7+ years across enterprise, crypto, and IoT. These days I'm doing design engineering at Quentic: building design systems and shipping AI-assisted features for B2B SaaS, and I built the design→dev workflow my team uses today (walked through below ↓).
How I think
- I bias toward shipping. A scrappy v1 in production beats a polished v2 in Figma.
- The strongest design calls are usually conceptual, not visual. The most important thing I designed for SDS Extract wasn't a screen. It was refusing to build a PDF viewer.
- Design systems are infrastructure, not artifacts. I treat them like a product: versioned, tokenized, with a roadmap.
- AI is a design problem with new rules, not a feature category. I'd rather write the manifesto than the style guide.
- My best work is cross-functional. I'd rather pair with an engineer or a PM to ship a fix than write a spec doc about it.
Tools I actually use
A new design→dev workflow
How I ship UX upgrades with AI.
Small UX upgrades (tooltips, alignment, microcopy, accessibility fixes) used to sit in the backlog waiting for engineering capacity. I built a loop that lets me ship them myself, then rolled it out as a workflow for the broader design team.
- 01
Design audit
I find the gap or rough edge, sometimes from user feedback, sometimes from my own crit.
- 02
Ticket
Written up like any product issue, with screenshots and acceptance criteria.
- 03
Implement with Claude Code
In our local env, against our Storybook, using the team's component library.
- 04
Open MR
Goes through normal code review with engineering, same bar as any other MR.
- 05
Iterate & ship
Address feedback, merge, deploy. The loop closes faster than a backlog ticket ever would.