Self-Initiated · Strategic Design 2025 – Present 🔧 Live Work in Progress

Global Adaptive
Design System

A strategic-design thesis disguised as a design system. Built for the 70% of internet users most products quietly ignore — non-Latin scripts, RTL languages, vertical reading directions, regional device tiers. The component library is the artifact. The argument is that direction-agnosticism is a market-access decision, not a localisation feature.

Type Self-Initiated Project
Status Work in Progress · Built Openly
Domain Design Systems · Global UX
Duration 2025 – Present
Tools Figma · Documentation
Why This Project Exists

Most design systems carry
three hidden assumptions

The way most companies scale globally today is broken. They hire regional design teams. They maintain parallel codebases. They ship "localised" products that are really just translated versions of a Western default. And even after all that effort, a massive portion of their potential audience still gets a second-class experience.

Here's the part most design systems don't acknowledge:

Assumption 01
660M
Users read left-to-right
People read right-to-left — Arabic, Hebrew, Urdu, Farsi. That's more than the entire population of Europe. Most design systems ignore them by default.
Assumption 02
40%
Users hold high-end devices
Of smartphones sold in Africa cost under $100. Another 40% under $200. These devices are not edge cases — they are the market.
Assumption 03
18%
English is the default
English is the first language of only ~18% of internet users, yet it dominates ~50% of online content. That gap is a product failure at scale.

This isn't a localisation problem. It's not a translation problem. It's a design systems problem. The frameworks most teams build are structurally incapable of adapting. GADS treats global adaptability as a first-class design constraint — not a post-launch patch.

How the System Works

Two independent axes.
Nine possible states.

GADS is organised around two independent axes. Any component in the system can be specified against any combination of them. The axes are independent and combinable — a component exists in up to nine possible states, and the system handles all of them without separate codebases, separate teams, or separate design files.

Axis 01 · Direction
Script direction awareness
LTR
Left-to-Right — English, Hindi, most European languages
RTL
Right-to-Left — Arabic, Hebrew, Urdu, Farsi
TTB
Top-to-Bottom — Japanese, Chinese, Mongolian vertical scripts
Axis 02 · Device Constraint
Three performance tiers
L1
Full — Flagship devices, stable 4G/5G. Full animations, rich visuals, spring physics.
L2
Reduced — Mid-range devices, 3G. Compositor-only animations, lighter treatments, 30fps acceptable.
L3
Minimal — Budget devices, 2G or metered. Zero animations, flat layouts, maximum performance. The product still works. Fully.
L1 — Full
L2 — Reduced
L3 — Minimal
LTR
LTR × L1🔧 In progress
LTR × L2⬜ Upcoming
LTR × L3⬜ Upcoming
RTL
RTL × L1⬜ Upcoming
RTL × L2⬜ Upcoming
RTL × L3⬜ Upcoming
TTB
TTB × L1⬜ Upcoming
TTB × L2⬜ Upcoming
TTB × L3⬜ Upcoming

Layouts aren't "LTR with RTL support." They're direction-agnostic from the ground up. The system replaces physical coordinates (Left, Right, Top, Bottom) with logical ones (Inline Start, Inline End, Block Start, Block End) — so the same component flows correctly in any script direction by design, not by override.

Demonstration Strategy

Stress-tested in
two product contexts

Two product contexts are being used to stress-test the system — one B2C, one B2B — chosen for maximum contrast. Every component in the system earns its place by appearing in at least two products or three screens, the way mature design systems actually emerge in practice.

B2C — Food Delivery App

Inspired by Zomato and Swiggy. High information density, complex flows, and a genuinely global use case. Screens: Home/Listing, Restaurant Detail, Cart & Checkout. Each adapted across all three directions and all three constraint levels.

B2B — Dashboard System

Inspired by Jira-style tools. Enterprise layouts, data-heavy views, and navigation patterns that most existing products already struggle with in RTL. The B2B context surfaces failure modes the B2C product rarely exposes.

The build methodology is bottom-up, not top-down. Wireframes first. Components extracted from real screen needs. Tokens and rules documented last.

What a System Like This Is Worth

Predicted outcomes,
grounded in benchmarks

Design systems have measurable ROI. A production-grade GADS-style system would likely deliver outcomes along these lines, based on industry benchmarks.

For the business

30–50% reduction in regional rollout effort Design and engineering cost per new market expansion. Sparkbox's Carbon Design System study found 47% faster build times for teams using a mature system.
247% cumulative ROI over 5 years A documented figure for mature design systems — Airbnb's DLS case study. The compounding returns come from reduced redundancy, faster iteration, and lower defect rates.
~25% of design team capacity freed Redirected from redundant regional execution toward strategic work — based on IBM Carbon programme data.

For the market

Access to ~660M RTL-reading users Without building a parallel product. MENA, South Asia, and diaspora markets are currently served by forked codebases or ignored entirely.
Access to ~1.5B low-end device users The users that flagship apps are currently unusable on. Southeast Asia's digital economy alone crossed $263B GMV in 2024 and continues to grow double-digit.

For the user

A product that functions on a 2019 Android device in Casablanca, in Arabic, over 3G. Visual and interaction consistency regardless of language, script, or device. No degraded "lite version" stigma — just the same product, correctly adapted.

Where Things Stand

Live work
in progress

This is a live project — built openly as part of my portfolio. The strategic framework is complete. Execution is underway.

Completed
✅
Strategic framing and business case — the problem, the axes, the benchmarks.
✅
System architecture — direction axis and constraint axis, fully defined.
✅
Component hierarchy and prioritisation framework — Tier 1 components identified across both product contexts.
✅
Research across production systems — Material Design, Ant Design, Cloudscape, Wikimedia Codex.
In Progress
🔧
Wireframe execution in Figma — B2C food delivery, LTR baseline first.
🔧
Tier 1 component extraction and spec documentation — components earning their place in the system from real screen needs.
🔧
Figma variable structure — Localisation modes and Performance modes as first-class Figma variables.
Coming Next
⬜
RTL and TTB variants across all wireframes.
⬜
L1 → L2 → L3 constraint-level transitions for every Tier 1 component.
⬜
After Effects animation showing the system degrading gracefully across constraint levels in real time.
⬜
B2B dashboard screens — enterprise layouts in all three directions.
Design systems are usually built to make life easier for the teams shipping products. GADS is built to make life easier for the people using them.

Want to dig into the thinking?

I'm building GADS openly because the strategic framework matters more than any individual component. If you're working on global product strategy, design-led transformation, or scaling design systems across markets — I'd love to talk. Currently exploring design-led digital transformation roles at Accenture Song, Deloitte Digital, and Publicis Sapient, plus UX strategy and digital transformation roles in Dubai.

Contact Mohit →