CORE: Design System & AI-Powered Figma Plugin

Founded a design system product and built an AI-powered Figma plugin from scratch. Managed a designer, led 8 developers, and scored 9.2/10 in usability testing with real designers.

Context

After building and maintaining design systems across 4+ companies, I kept running into the same problem: systems that looked great in Figma but fell apart in production. Tokens drifted, components diverged, and every new brand meant hours of manual retheming. I decided to build the tool I wished existed.

Problem

Designers build components in Figma. Developers rebuild them in code. Over time, the two drift apart -colors don't match, spacing breaks, and every brand customization becomes a manual, error-prone process. Existing tools treated this as a documentation problem. The real issue was that theming and token management were still manual workflows with no automation layer.

My role

I owned everything: product vision, design, and development. I designed the full component library with semantic token architecture, then built the Figma plugin myself using Claude and Cursor. I collaborated with a developer on plugin v1, managed a product designer on system adoption and contribution standards, and stepped in to lead a team of 8 developers for two months during a team lead transition.

Key decisions

Built the plugin myself using AI-assisted coding (Claude + Cursor)

Why: No engineering team was available, and the plugin was the core value proposition -I couldn't wait

Trade-off: Steep learning curve, but shipped a working product in weeks instead of months

Ran usability testing before public launch

Why: Needed real validation, not assumptions -the system had to work for designers outside my own workflow

Trade-off: Delayed launch by two weeks, but the feedback directly shaped the final product

Designed for mobile-first despite web demand

Why: Mobile components are harder to get right and the market was underserved -web could come later

Trade-off: Multiple testers tried CORE for web anyway, confirming web as the next roadmap priority

Testing

Structured usability test across experience levels. Each tester got the beta file, designed a mobile screen using CORE, then filled out a detailed feedback form.

"Everything was perfectly categorized and easy to navigate. And the important one - all layers were named." (Mid-level product designer)

Testers flagged limited icons, input size gaps, and typography weight variations. Multiple testers tried CORE for web even though it's mobile-first - a signal that directly shaped the roadmap.

Outcome

CORE shipped as a 40+ component mobile-first design system with an AI-powered Figma plugin. Usability testing scored 9.2/10 across 10 designers of varying experience levels. The plugin cuts brand color theming from 30 minutes to under 1 minute. The project attracted 900+ community users and has been battle-tested across fitness, longevity, fintech, and SaaS products in Toronto and Dubai.

Why it mattered

CORE wasn't a side project -it was a full product bet. I identified a real gap, built it from zero, shipped it, and proved it worked with real users and real data. Along the way, I managed a designer, led developers, and made every product decision. It's the clearest proof of what I do: take a complex problem and turn it into something that works.

Links