Case Study — Imagio

Reimagining jewellery photography with AI.

Imagio is an AI-powered photography studio that lets jewellery brands create new product shots without organising another photoshoot. Upload one photograph, then change the shoot as easily as you’d change a setting.

AngleLightingLensPoseBackgroundPresentation

Role

Product Design · UX ·
Product Build

Platform

Web

Stack

Next.js · TypeScript · Three.js ·
Supabase · Polar · OpenAI

Type

0→1 AI product, built end to end

Scroll

02 — The problem

Jewellery photography is expensive and repetitive.

A single product may need multiple angles, lighting setups, lenses, poses and backgrounds—often requiring new shoots, new models and fresh production time for what is, essentially, the same piece seen differently.

So I wanted to explore a simple question:

What if changing a photoshoot was as easy as changing a setting?

03 — The idea

A digital photography studio, not a prompt box.

Instead of treating generative AI as a text field to wish into, I designed Imagio as a studio. A user uploads an existing jewellery photograph and experiments with the same variables a real shoot would—except every dial is now immediate.

ControlCamera angle
ControlLighting
ControlLens
ControlPose
ControlBackground
ControlPresentation

The AI generates a new version, but the studio is what gives the user real authorship over the creative direction—rather than the lottery of a single prompt.

04 — Designing the studio

Familiar photography concepts, generative underneath.

A 3D camera and lighting viewport helps users understand and manipulate the scene the way a photographer thinks about it—where the camera sits, where the light falls. Presets, reusable poses and recipes then make repeatable workflows possible, so a look developed once can be applied across a catalogue.

Around the creative core, the studio handles the practical parts of an actual product:

Projects Upload Configure Generate Compare Save

Generations are stored in history, projects auto-save, and a credit-based system supports different subscription tiers—the unglamorous scaffolding that separates a demo from something a brand can actually rely on.

05 — Building the product

Beyond a prototype, into a working product.

I took Imagio past a visual prototype and built the working product with Next.js, TypeScript, Three.js, Supabase, Polar and OpenAI’s image-generation model. That meant designing for the states a real generative product lives in—not just the happy path.

Generation states

Queued, generating, ready—the wait is part of the experience, not an error screen.

Failed generations

The model doesn’t always cooperate; the product has to fail gracefully and keep credits honest.

Credits & tiers

A credit system that maps cleanly onto subscription tiers and real generation cost.

Storage & history

Every generation kept and retrievable, so iteration has a memory.

Auth & persistence

Accounts, and projects that auto-save so nothing creative is lost between sessions.

Studio architecture

The underlying model that makes angle, light and lens composable rather than hard-coded.

06 — The takeaway

What happens when AI becomes the studio itself?

Imagio explores what happens when the model isn’t just another tool inside the studio—it becomes the studio. The goal was to move away from “write a prompt and hope for the result” toward a familiar creative workflow where designers can direct, experiment and iterate.

“Direct the shoot. Don’t gamble on a sentence.”
Want a walkthrough of the studio? Book a call