Role

Product Design Manager

Scope

Accelerate Product Development with AI

Brands

Junglee Games & Raffles

Tools

Figma, Claude, Gemini

AI-Enabled Product Infrastructure for Multi-Brand Platforms

Building Scalable UI Infrastructure to Accelerate Product Development with AI

Overview

At Junglee Games, multiple products operated across different brands, geographies, and betting formats — including poker, raffles, and other gaming experiences. As product complexity increased, the speed of feature rollout and experimentation started slowing down due to fragmented UI structures and duplicated effort across teams.

I initiated and proposed a strategic infrastructure initiative focused on creating an AI-ready

shared product UI infrastructure that would:

• Standardize foundational UI architecture
• Support multiple brands and markets
• Enable AI-assisted product development
• Improve product velocity across teams

The goal was not just consistency — it was to create a scalable infrastructure layer that could significantly accelerate future product creation using AI tools like Claude and Gemini.

Problem

As the organization scaled, several operational bottlenecks emerged:

Fragmented Product Structures

Each product evolved independently with its own:
● UI patterns
● components
● layouts
● visual structures
This created duplication and slowed feature development.

Multi-Brand Complexity

Different gaming products required:
● unique branding
● market-specific customization
● localized UX patterns
Scaling across brands became increasingly inefficient.

Slower Product Velocity

Feature development involved repetitive manual effort:
● recreating screens
● redesigning patterns
● rebuilding layouts
This increased both design and engineering overhead.

AI Couldn’t Be Leveraged Effectively

Modern AI tools require structured systems to generate reliable outputs. Because the product ecosystem lacked unified infrastructure, AI-generated UI workflows were inconsistent and difficult to scale.

Opportunity

I identified an opportunity to treat UI architecture as shared infrastructure, similar to platform engineering.

The vision was: Build AI-ready product infrastructure that allows AI tools to accelerate product development across multiple brands and products.

Instead of designing each product independently,the system would separate:

● structure
● branding
● templates
● product logic

This would allow both humans and AI systems to build products significantly faster.

Approach

Infrastructure-First Thinking

Rather than positioning this as a “design system,” I reframed it internally as:

Product UI Infrastructure

This aligned the initiative more closely with:

● scalability
● execution speed
● platform architecture
● operational efficiency

This aligned the initiative more closely with:

System Architecture

The infrastructure was structured into multiple scalable layers:

Multi-Brand Infrastructure

The system was designed to support multiple gaming brands using a shared foundation. Instead of duplicating components for each brand, I implemented:

Shared Components

Reusable components across all products:

● navigation
● cards
● betting modules
● onboarding patterns
● transactional flows

Instead of hardcoding values like 4px or 16px, spacing tokens are derived from a
consistent scale.
This ensures proportional rhythm across the entire system.

Brand Modes

Brand-specific visual identities were controlled through variable modes:

● Poker
● Raffles
● Country variants
● Future products

This allowed one shared structure to adapt visually without rebuilding screens.

Layout & Template System

To improve scalability and AI generation reliability, I introduced two critical abstraction layers:

Layouts

Structural skeletons defining:

● content hierarchy
● scrolling behavior
● navigation structure
● dashboard patterns

Example:

Layout/Mobile/Dashboard

Layout/Mobile/List

Layout/Mobile/Checkout

Templates

Product-aware reusable screen blueprints.

Example:

Template/Home/Poker

Template/Home/Raffles

Template/Onboarding/Compliance

Templates combined:

● layouts
● components
● business-specific patterns

This dramatically accelerated feature creation.

AI Integration Strategy

The infrastructure was intentionally designed to support AI-assisted workflows.

Using Claude

Claude was used primarily for:

● structured UI generation
● layout hierarchy creation
● template generation
● multi-variant exploration
● system-aware screen creation

Claude performed especially well when constrained by:

● predefined layouts
● semantic naming conventions
● reusable component architecture

Using Gemini

Gemini was leveraged for:

● rapid ideation
● flow exploration
● UX variation generation
● product scenario simulations
● content and interaction modeling

Gemini became highly effective once the infrastructure provided enough structural clarity.

Why Both AI Systems Were Valuable

Claude and Gemini served complementary roles:

Claude

Structured generation 

System discipline 

Hierarchical UI creation 

Template consistency 

Gemini

Fast ideation

Exploratory thinking

Product brainstorming

Flow experimentation

The real multiplier was not the AI model itself — it was the infrastructure enabling both systems to operate effectively.

Outcome

The initiative established the foundation for:

Faster Product Development

Shared infrastructure significantly reduced repetitive design effort.

Multi-Brand Scalability

New brands and market variants could be introduced without rebuilding UI foundations.

AI-Assisted Workflows

Teams could begin leveraging AI to:

● generate screens faster
● explore feature variants
● accelerate experimentation

Better Long-Term Maintainability

Separating:

● structure
● branding
● templates
● product logic

made the ecosystem significantly easier to scale.

Key Learning

The biggest realization from this project was:

AI does not replace product systems.

AI amplifies the value of structured systems.

Without infrastructure, AI output becomes inconsistent.

With infrastructure, AI becomes a force multiplier for product development speed and scalability.

Reflection

This project shifted my perspective from:

designing screens

to:

designing scalable product infrastructure

It reinforced the idea that the future of product design is not just visual consistency — but building systems that enable faster creation, experimentation, and scaling through AI-assisted workflows.