Overview
I was brought in during the desktop alpha to improve the existing UX/UI and adapt the game for mobile. I started by cataloguing every UI element in the build, converting experimental files into proper Figma components, and auditing the flows to find the ones with the sharpest user issues. From there I fielded feedback from QA, the internal team, and alpha desktop testers, working with product managers, designers, and developers to decide what needed fixing for the mobile alpha versus what could land after launch.
Core feature redesign covered a brand-new Food Quota UX/UI for the feeding mechanic, a complete revision of the astronaut market, and clearer interfaces for selling and assigning food so those actions connected to the other core mechanics. Then came the full mobile adaptation — turning “big windows” designed for desktop into phone-legible interfaces, swapping drag-and-drop and hover states for touch and touch-hold patterns, and using visual indicators to keep actions clear. I then re-aligned desktop to match the new mobile system so both platforms felt like the same game. On top of that I designed UX/UI for new post-level-20 features (mole mining, microwave production buildings, more) and produced icons and spritesheets — a mix of AI artwork and Photoshop — to help developers with optimisation.
Outcome
Starvin’ Martian launched its mobile alpha in late 2025 with markedly improved UX/UI on both platforms. QA and internal testing confirmed clearer interactions and better hierarchy, and the design system aligned mobile and desktop around the same visual language. Team reception was strong, with the critical fixes landing inside an aggressive two-month window.
What I learned
Joining a game mid-build is a lesson in restraint. The temptation is to redesign everything; the right move is to triage, fix what hurts the player most, and only then move to system work. The cross-platform pass reinforced something I kept running into at Illuvium: translating drag-and-hover interactions into touch isn’t a shortcut, it’s a full re-design of that interaction — and the only way to keep consistency across platforms is to stop treating one as a variant of the other.