Behind the Counter

Reimagining Mobile Education Through an Immersive Web Experience

Behind the Counter?

Haneke Design partnered with a national public safety organization to create a mobile-first immersive web experience for law enforcement education. Designed for convention attendees and kiosk deployment, the React-based application contrasted the bright façade of illicit vape shops with the criminal activity often uncovered behind the counter.

The platform uses a custom React-based CMS with a Supabase backend to enable real-time content updates and dynamic hotspot management. Design assets were created using ChatGPT DALL·E, Firefly, Adobe Photoshop (Nano Banana), and Figma with Figma Make (Claude). The interactive 3D experience was developed with Claude Sonnet 4.0, Three.js, and WebGL.

The Challenge

Teach quickly. Connect deeply. Respect people’s time.

Our client came to us with a clear goal: build a mobile experience for law enforcement convention attendees that reveals the hidden risks in vape and smoke shops. It needed to work on people’s phones through a QR scan and run smoothly on event kiosks, too.

Most importantly, it had to show a real contrast: what looks fine up front vs. what investigators sometimes find hidden in the back.

A static PDF just wouldn’t cut it.

Constraints

  • Loads instantly, no hassle
  • No more than three sentences per insight
  • Clear source attribution under every claim
  • Big, easy-to-tap buttons
  • Mobile-first performance

The Concept

“Behind the Counter”
The experience unfolds in two acts.

The counterBehind the counter

Scene One

A brightly lit vape shop interior. Candy-colored packaging. Youth-targeted signage. A storefront that feels typical. Six tappable hotspots reveal concise, sourced insights, from unauthorized brands to misleading “FDA Approved” signage to flavors banned domestically in China but exported to the U.S.

Then the invitation appears:
Go behind the counter.

Scene Two

Lighting drops. Colors desaturate. Shelves clutter. Hidden compartments emerge. The narrative references real enforcement actions, including DEA-led initiatives such as Operation Vape Trail.
Ten hotspots expose the deeper network: concealed inventory, paraphernalia, cash, cartel distribution pipelines, and enforcement action.

The shift isn’t just informational.
It’s tonal.
Perception gives way to investigation.

Choosing the Right Platform

Why This Became a React Application

Website builder tools fell short in three critical areas

  • Animation control
  • Scene layering and depth
  • Kiosk reliability at scale


We picked React so we could build this like an app, not just a website.

Building in React gave us fine-tuned control over

  • Component-based hotspot systems
  • Dynamic state-driven scene transitions
  • Modular popup architecture
  • Flexible positioning logic for interactive elements
  • Conditional behavior for kiosk vs. personal devices
  • Performance optimization for layered visual assets

This wasn’t just a scrolling story. It was an environment you could actually interact with.

Beyond the Front End

A Custom CMS for Real-Time Updates

Because enforcement data evolves, the client needed the ability to update hotspot content without redeploying code.

We built a secure, client-level CMS panel directly in React.

Administrators can

  • Edit hotspot copy
  • Update source links
  • Adjust hotspot positioning coordinates
  • Toggle visibility
  • Manage scene-specific content

On the backend, we implemented Supabase to manage​

  • Authentication and role-based access
  • Structured hotspot content
  • Positioning metadata
  • Source attribution data

This architecture ensures the experience remains current as enforcement briefings evolve, while preserving performance and security.
A flexible, modern backend powers the immersive front end.

Building the World

AI-Generated Imagery, Human-Refined Craft

We faced a real visual challenge: how to craft believable environments without imitating actual cases or locations.

AI tools were used for environment creation and scene refinement

  • ChatGPT Dali for overall scene generation and prompt development, which produced the strongest results for overall environment design
  • Adobe Firefly for sections in the scene visuals and major section or scene edits

No single tool could do it all

Each image was composited, layered, color-graded, and manually refined in Adobe Photoshop with the Nano Banana model for object generation and detailed scene polish. Scene one pops with neon brightness, while scene two shifts to muted, gritty tones—a palette change that became part of the story itself.

Creating Depth in the Browser

CSS, WebGL, and Subtle Cinematics

Flat imagery would have reduced impact. We needed dimensionality without sacrificing mobile performance.

Where appropriate, WebGL and Three.js were integrated to enhance dimensional realism while maintaining load performance. The end result feels cinematic even though it all happens right in the browser.

Creating Depth & Atmosphere

  • Layered parallax effects
  • CSS blend modes and masking
  • Environmental light gradients
  • Shadow overlays for atmospheric depth
  • Smooth animated scene transitions

AI-Assisted Production

Accelerating Experimentation Without Sacrificing Craft

This project doubled as a forward-looking production experiment.

Design Phase

  • Core UI and workflows developed in Figma
  • Rapid functional prototypes explored using Figma Make
  • Interaction patterns validated before development

Development Phase

  • React implementation accelerated with Claude Sonnet 4.0
  • AI-assisted CSS experimentation allowed faster animation refinement
  • Iterative UI adjustments tested live, reducing feedback cycles

AI didn’t replace our expertise. It compressed iteration.

Designers moved fluidly between visual concept and functional prototype. Developers could test animation ideas in minutes instead of hours. The boundary between design and engineering narrowed.

Mobile-First by Discipline, Not Just Declaration

Every UX rule was intentional

44px minimum tappable areas

Maximum three sentences per hotspot

One-line source attribution under each claim

No gating before exploration

Clear exit to download the full enforcement brief

The experience respects the audience’s time.

It invites discovery rather than demanding it.

What This Project Proves

Design Smarter with AI

How AI tools help designers save time and focus on creativity.

brainstorm
concepts

Generate
mockups

suggest layouts,
compositions

speeds up
repetitive tasks

With AI handling time-consuming tasks, designers can focus on strategy and creativity.

Result = More Time. Better Ideas. Faster Design.

  • Mobile web can deliver immersive, app-like narrative without installation.
  • AI-assisted image generation can dramatically compress environmental design timelines.
  • AI-assisted coding enables faster visual experimentation.
  • React combined with Supabase creates a powerful foundation for dynamic experiential platforms.
  • Public-sector education benefits from thoughtful design systems, not just content accuracy.
  • Regulated industries and public safety organizations are often underserved by immersive digital experiences. When clarity, credibility, and craft must coexist, interactive web becomes a powerful medium.

Let’s Build What People Don’t Expect

Whether you’re teaching, informing, or changing minds, immersive mobile experiences create moments you just can’t get from static content.

If you’re exploring new ways to engage audiences, especially in complex, regulated, or public-sector environments, let’s talk.

Additional Case Studies

Movotic Care Assistant

A proactive patient support platform designed to simplify healthcare communication

Think Big for Kids

A platform to connect students with mentors, career resources, and growth opportunities

One World Observatory

An immersive experience unique to the Big Apple

Delete before going live this is the hover card

Static

VS animated

Behind the counter