Project Overview

Role: Lead Product Designer (Sole designer)

Team: Founder + 3 Engineers + 2 Growth Managers

Timeline: 6 weeks, 2025

Scope: Brand identity, product design, design system, social, landing page

Outcome: 100M+ Total Volume, 10k+ Swaps across MegaETH Testnet

The Challenge

MegaETH chain is built around real-time execution, <10ms confirmations on an L2 that's positioned to host interfaces that finally feel like centralized exchanges. But every existing DeFi UI was designed for chains where confirmations take minutes. A user who wants to swap, then LP, then borrow against that LP, navigates three apps, three wallet rituals, three visual languages. And ultimately had to trust three interfaces while liquidity remains trapped in silos.

The Solution

A single superapp where five financial primitives: trade, swap, lend, borrow & LP share one shell, one data model, and one interaction language. The order book leads, rather than the swap, so the chain's real-time advantage is visible the moment a user arrives. A retro-futuristic brand language set against a Bloomberg-density product UI, designed so that a degen launching a token and a first-time swapper can use the same surface without either feeling out of place.

Brand identity

The brief was a contradiction on its face: the product is bleeding-edge infrastructure, but bleeding-edge infrastructure brands all look the same i.e., saturated gradients, geometric sans serifs, the same five purples. We wanted something that felt primitive and warmer than that (Retro-futuristic). Trading terminals from the past, but running on a chain that confirms in milliseconds.

That tension drove every decision. The palette went deep oxblood instead of the usual crypto purple. The product UI stays cool and dark for legibility, but the brand surfaces i.e. landing, social, hero moments, lean into the warm, retro register.

The logo went through dozens of iterations. The R was the obvious starting point, but it took a long time to find a form that felt right. Early versions were too clean, they could have been any tech company. The breakthrough was treating speed itself as a brand element. In a real-time exchange, the millisecond is the product. So the final mark carries a flash of a literal speed cue baked into the letterform built on a custom letterform we drew specifically for the logo, not picked from a typeface.

Some logo & brand birection explorations:

Landing page

The landing page is the brand's strongest single statement. After several rounds of iteration, the final concept centers on a retro Macintosh frame housing a live trading terminal, lit with hard light rays. The image does the argument for you: this is real-time finance, but it remembers where finance came from.

Rejected Landing Pages:

Approved Landing Page:

Social kit

The social system was designed as a first-class part of the brand, not an export from the product file: a full set of Twitter headers, profile assets, and post templates built to make the timeline feel like a single coherent presence rather than a collection of one-offs.

The product

Five surfaces, one product. The brief was less about inventing new patterns and more about making five financial primitives feel like a single coherent thing, same shell, same data grid, same interaction model, retuned per surface.

Swap

The Swap surface uses the same data grid and shell as trade, so a user moving from the order book to a quick swap doesn't feel like they've changed apps. The token selector, route preview, and confirmation states are tuned to feel fast; every transition is a deliberate motion beat rather than a hard cut, so the interaction reads as continuous even across multiple screens of state.

Trade

The trade is the centerpiece because it's where the MegaETH chain's advantage is most visible.

I started dense, hyperliquid as the explicit reference point because it's the cleanest implementation of pro-trader density in DeFi right now, and pulled back from there. The first pass tried to fit too much in a single screen. Got team feedback, tightened it: the data is dense, but the priority hierarchy isn't. The order book, the chart, the position panel, and the order form each had to claim a tier of attention without competing.

Earn

The liquidity pools are split between v2 & v3 depending upon the type of fee in the pool. Gave a smooth flow to the users for providing liquidity unlike tens of screens.

Launchpad

Built for the degens hunting new tokens at the moment of launch. The constraint was specific: dumping 100 tokens into a single list creates overwhelm and obscures the only thing a degen actually wants to know i.e. the token lifecycle.

So Launchpad splits into three explicit sections: newly launched, graduating, and recently graduated. Same data grid, three contexts. A user who wants to snipe an early entry scans Newly Launched. A user who wants to trade momentum on a token about to hit a major listing scans Graduating. A user who missed the early window scans Recently Graduated. Risk appetite maps directly to the section.

Lend & Borrow

This was the hardest part of the project, and to be honest, I'd never designed a lending product before.

I treated it as a research problem. Tested some competitors like Aave for the canonical patterns & Morpho for the modern density take. Then, pressure-tested every screen with the founder, who'd shipped this domain before and could spot the small mistakes and the kind of details a first-time lend/borrow designer wouldn't know to flag. The screens that came out of that process are the ones I'd most want a senior reviewer to grade me on.

Lend and Borrow were fully designed in Figma. They haven't shipped in production as the engineering scope didn't fit in the 6-week window.

Outcome

Realtime Exchange shipped on testnet ahead of MegaETH mainnet going live. Within the testnet window:

• $100M+ in trading volume

• 100,000+ transactions

• Thousands of unique users and token launches