ENFR

Work // Pump

Same architecture, new identity, and a system to hold it together.

Jan–May 2026Collaboration w/ Juliana Ponze

Background

Pump is a sister platform to Bitsler, a crypto casino. Same underlying interface, different audience: younger, European, and slightly opposed to the crowd Bitsler was built for.

Problem

A new brand had to feel like its own thing while sitting on an interface that already belonged to someone else.

Approach

Identity first, then a design system built on states rather than components, because green already meant three different things.

Outcome

A full brand identity, a design system, and a redesigned interface from sign up to sign out.

72

Component sets

209

Screens

22

Weeks to ship

Case study

Brief

A new brand that has to wear its sibling's interface.

Pump is Bitsler's sister brand, aimed at a younger European audience: fast, bold, privacy-coin adjacent. It shares the same UI base as Bitsler, which had just come out of a redesign. The identity had to feel like its own thing while sitting on an interface that already belonged to someone else.

Constraints

Three things shaped every decision.

  • A borrowed interface. Pump inherits Bitsler's UI base, recently redesigned. Differentiation had to come from identity, not layout.
  • A moving deadline. Six month original deadline for publish, license delay.
  • Colour locked early. The green and pink direction was approved as soon as it was shown, which set the palette before the exploration was finished.

Exploration

Two directions, one shared interface.

Five directions went up. Two came back. Both had to work on an interface that was already built, so the difference had to live in the identity: the mark, the palette, the language, the tone.
Both directions tested on the same screens before either was chosen

Direction 1

Electric Hare, catch the charge, ride the pump.

  • Fast, arcade energy. The hare is speed, the bolt is the pump.
  • Game menus and neon nightlife, playful, aimed younger.
  • Unique colour combination, and close enough to Bitsler's language to sit alongside it.

Direction 2

Captain Pump, raid the odds, feed the troop.

  • Chest-thumping treasure hunt. Missions, raids, chests, XP races.
  • Loudest and most character-driven of the two.
  • Gorilla, pirate, treasure, raids, XP races.
  • A disruptor in the category, with room for language and play across the identity.

Neither won outright. The green and pink palette from Direction 1 was an instant yes, so that stayed. What carried Direction 2 was the ape: what it stood for, and how much play the name gave us across the identity. The final brand is a blend, the palette from one and the character and language from the other.

Identity

Confident, sharp, playful, premium.

Hellix for headings, Lexend for everything else. Green for action and gain, pink for identity and secondary emphasis, greys doing the structural work. The condensed sheet is what everyone else worked from: personality, values, tone of voice, do and don't. Short enough to be read, specific enough to be followed.
The condensed identity, written for the people who had to apply it without me

Style

Soft glow and 3D, or flat vector.

Two treatments of the same screens. Style 1 leaned on depth, glow and cursor-reactive surfaces. Style 2 stayed flat, with rounded soft-edged buttons and vector artwork. Style 2 won, then got pushed further away from realism than the original thumbnails suggested. Less rendered, more graphic, and easier to hold consistent across dozens of assets.
Same screens, two treatments, the flat one shipped

Applied

Thumbnails are where the style gets tested.

Small, dense, side by side, and each one still has to be identifiable at a glance. This is where the flat direction was proven or not. Promotional material sat on one template with a written material rule: surface steps, subtle gradients, a grain overlay, no outer glows, no pulsing. That rule is what let marketing produce assets without the brand drifting.
Thumbnails, theme cards and promo built on the same material rule

Design system

Pump is green. Winning is green. The primary action is green. On a dark interface that stops working fast.

Green is earned by the moment, not assigned to the component. A button lights when it is the thing to do right now and goes quiet when it is not, so the interface only ever carries one or two live greens. Nothing is permanently primary.

States are where the system either holds or falls apart. Every component was defined by what is possible at that moment, not by what looks balanced on the board.

State definitions from the file, including the notes other designers worked from

Overview

Same interface, two brands.

Bitsler on the left, Pump on the right, one layout carrying two identities

Side quest

New platform, new welcome offer, so nothing had to inherit anything.

The structure was complex: three deposits, an XP ladder, free spins and wagering conditions, and most of it is invisible until someone has already committed money to it. I mapped the first session end to end to find where someone stops knowing where they stand. Every one of those points became a moment the interface has to report back: what was earned, what it is worth, what unlocks next.
Journey map of the first session, each highlighted node is a point where the interface has to report back

Two paths, because the offer suits different play. Instant credits the bonus on deposit. Unlock releases it in batches against wagering milestones. Both state the withdrawal condition up front rather than behind a terms link.

It fell outside the delivery window and did not ship.

Outcome

12 game types running on one action system.

72 component sets, 209 screens, two designers on UI, 22 weeks from first direction to handoff. Brand, design system and welcome offer logic built alongside a live rebuild of the sister product.

What I would change: the colour was locked before the exploration was finished. Next time I would push back and keep the palette open longer, because making two strong colours work across a dense dark interface is a different problem to making them work on a logo. I would also have pushed harder for the promotional and campaign side, where the brand had the most room to play.