← ZAR · Case study · 2024 – 2025
670+ components, 37 Figma pages, 103 icons, 3 platforms.
ZAR's product was freelancer-built — multiple designers, multiple engineers, no shared system. I built Techtonic three times as the product pivoted underneath it. Each version carried forward lessons; the third shipped and held. The substrate that made 110-country coverage possible without 110 engineering teams.
ZAR Techtonic · Design System · 2024 – 2025
ZAR's Figma files were a graveyard of freelancer-built components, orphaned frames, and styles that varied by who designed them and when. Every screen looked slightly different. Three platforms — iOS, Android, web — each with their own inconsistencies.
The product itself was evolving rapidly. ZAR pivoted its positioning multiple times, which meant the design system couldn't be a static artifact — it had to keep pace with a moving target.
Version 1 was the cleanup — an audit of every component, every duplicate, every style inconsistency, consolidated into base tokens and core components. It brought order to chaos but was designed for a product scope that soon expanded. Version 2 introduced platform-specific adaptations: each platform got its own component set sharing the same tokens but respecting platform conventions. V2 also introduced the icon library — 103 custom icons designed specifically for ZAR's product vocabulary.
Version 3 was the one that shipped. Integrated with GitHub and Storybook, engineers got a coded component library that mirrored Figma 1:1. Token architecture in three tiers: primitive, semantic, component. 37 organized Figma pages covering primitives, patterns, templates, platform adaptations.
Each version incorporated lessons from the last. Battle-tested against real product changes.
Techtonic, three iterations
Engineers could self-serve from Storybook. Design reviews shifted from "does this match the design?" to "does this solve the user problem?" — a much more valuable conversation. The system held through pivots because its architecture was designed to bend without breaking.
Techtonic became the substrate that made 110-country coverage possible without 110 engineering teams. Three iterations, one shipped system, and a team that now moves faster because the rails are already laid.