Work
About
All Projects

Case Study · 2024

AI Design System

Generative UI component ecosystem

Role

Design Systems Lead

Duration

6 months

Year

2024

Tools

Figma, Cursor, AI Studio, Storybook

Built a design system where components are generated and contextually adapted by AI based on brand context, content density, and user intent — reducing manual design effort by over half.

The Problem

Traditional design systems require constant manual maintenance and fail to adapt to diverse brand requirements at scale. Every new brand meant weeks of token overrides and component forking.

Process

01

Audit

Analysed 5 existing design systems. Catalogued the 80 most-duplicated patterns and identified the 12 highest-friction customisation points.

02

Grammar Design

Defined an AI generation grammar — a structured language for describing UI intent that the model could reliably translate into component props.

03

Token Architecture

Designed a 4-tier token system: global → semantic → component → override. Built automated token sync between Figma variables and code.

04

Documentation

Wrote component API docs, usage guidelines, and AI prompt templates. Ran internal workshops with 12 teams.

Project screens / mockups

Replace with images from your case study

Outcomes

−55%

Design Time

per new brand setup

200+

Components

in the library at launch

12

Teams Adopted

across 3 product lines

Previous

Benchmarks.Digital

Next

Mobile App

Anand
Gaurav

Get yourself a laptop, please.

Contact