I joined to design a staking app. 4.5 years later I owned the design of every surface the company shipped,
and the team that made it.
Role
Senior Designer, then Lead, then Head of Design.
Scope
Brand, game UX/UI, product, marketing
Team
Grew from 2 to 5 Designers
Shipped
4 games, launcher, marketplace, sites
Illuvium was building 4 games at once. An open-world creature collector, an auto-battler, a NFT collection game, and a city builder title. Different genres, different art directions, different teams inside the same company.
All of them had to feel like Illuvium, all of them had to be playable by people who, for some, had never opened a crypto wallet.
Everything I did there was a version of that problem:
how do you hold a visual identity together across surfaces that have every reason to drift apart, without flattening what makes each one good.
01 ILV & sILV tokens
My first task at Illuvium was the ILV and sILV2 token logos. They became the company's primary symbol, used across socials, web, merchandise and every game surface that followed.
02 Starting with the hardest interface in the product
My first project was the Staking app. The least glamorous surface and the most revealing one: real money, irreversible actions. If someone misread a screen here, they lost funds.
Designed around one rule: never show a number without showing what it means and what happens next. This project set the vocabulary I would reuse for years. The way we handled loading, confirmation, error and empty states in the staking app became the baseline for everything else, including the games.
03 A Design System that four games could actually use
Game UI resists design systems. Each title has its own art direction, its own camera, its own reasons to break a grid. The Game Design System created is a structural layer, non-negotiable, covering spacing, type scale, interaction patterns, focus and input states.
First attempt
One component library, one visual language, applied top down. Game teams overrode it constantly. The system described what we wished were true rather than what shipped.
What worked
A shared structural spine with a per-game skin layer. Teams got real ownership of the surface, and players still moved between titles without relearning anything.
04 The launcher, where the whole ecosystem becomes one thing
The launcher was the only surface where all games sat next to each other. It was where the coherence problem stopped being theoretical.
It also carried a lot of weight beyond installing a game: account and wallet, asset ownership, news, updates, and the first thirty seconds of a new player's relationship with Illuvium. I designed it to lead with play and keep the Web3 machinery available but out of the way, so the on-ramp did not read as a finance product.
05 Brand at the speed of a live community
Alongside the product work I owned the brand as it was used publicly: launch moments, social assets, the marketing sites. In a game community the brand is not a guideline document, it is whatever people saw this week.
The volume was the constraint. I built templates and asset kits so the team could produce quickly without every post becoming a design review, and kept a small set of things genuinely fixed: logo behaviour, type, the way game art was framed.
06 Leading the team
Team size
2 to 5 designers
Time as lead
4 years
Surfaces owned
Games, launcher, web, brand
Working model
Fully remote, global
I moved from Senior Designer to Lead to Head of Design in 18 months, which mostly meant learning where to stop designing myself. I hired for range rather than specialism, because a small team covering four games cannot afford anyone who only does one thing.
What I held onto: reviewing every surface before it shipped, and staying hands-on with the brand. What I let go of: being the person who made the files. The design system existed so decisions did not have to route through me.
07 They asked me back
I left in October 2025. In December the founders brought me back to work on Illuvium Deathmatch for a few months, handling branding, art direction and product design for the new title through June 2026.
I include this because it is the only part of a case study a designer cannot write himself. Four years of work is a claim. Being asked to return is evidence.
