Square-AI

 

Go/Fake Figma Plugin

Designers spend more time hunting for content than designing. Go/fake Figma Plugin was built to fix that. Developed with Cursor and Square AI’s Goose, the plugin pulls fake merchant’s data from Coda and Notion databases directly into Figma — letting production designers fill screens with accurate, structured content at the speed of design.

Production Designer & Tool Developer

Role Pills
Systems Thinking UX/UI Design & Prototying AI Application Figma Plugin Development Internal Tooling

 


Overview

Go/Fake Figma Plugin is an internal Figma plugin built to streamline screen UI production at Square. It connects directly to Coda and Notion databases, letting designers pull structured content — copy, pricing, cart calculations, and imagery — straight into Figma compositions, eliminating the manual bottlenecks that slowed every campaign.


The Challenge

Screen UI production for marketing required building compositions across a large library of Figma screen components — each one manually populated with copy, imagery, and metadata sourced from external databases.

At scale, the process broke down quickly:

  • 11 languages across 8 locales per campaign

  • Content copied field by field with multiple databases from Coda and Notion into Figma

  • Images sourced manually from separate Figma libraries

  • Translator QA was required for each localized version, adding rounds to every campaign


The Opportunity

The rise of AI-assisted development tools like Cursor and Claude AI created a new possibility — production designers could now build and own their own tooling without relying on an engineering backlog.

I saw an opening to solve a problem my peers and I encountered every day, and built the prototype and solution to aid this time consuming work-stream.


Project Highlights

 

Square's product ecosystem spans a wide range of verticals — Point of Sale, Restaurants, Retail, and Appointments — each with its own extensive Figma component library.

Go/Fake Figma Plugin made populating these screens for production significantly faster, more accurate, and consistent across every product line.