Lebi · Creator and design lead · 2025
It began as scattered ideas. I gave it a spine.
A free sports predictions platform I created and led at MoveUp Media: players compete for real prizes, and brands launch sponsored leagues on the same template.
Shown in English with prizes in US dollars. The product runs in Portuguese, for Brazil.
- user interviews
- 20+
- personas mapped
- 4
- fewer design questions from engineering
- 40%
- screens in the light theme
- 81
Overview
Lebi is a free sports predictions platform. Players compete by predicting matches and win real prizes, and brands launch sponsored leagues on the same template. I created it and led its design at MoveUp Media, taking it from scattered ideas with no identity to a shipped SaaS: the positioning, a modular design system from day one, the research, and the calls that shaped the product.
Opportunities
Structure from chaos
How might we turn scattered ideas and no identity into a product with a clear spine, quickly?
A game, not a bet
How might we position Lebi as a game to play, not a betting product, and hold that line across the whole experience?
Newcomer and power user
How might we serve a curious newcomer and a data-hungry power user without overwhelming either?
Validate before building
How might we test game mechanics early, before engineering commits a sprint to them?
Prizes people can trust
How might we pay out real prizes safely without scaring away players who just want to play?
Any sponsor, same template
How might we let a brand launch its own league without new design work?
Approach
- 01
Positioning first: a game, not a bet
Before any screen, I fixed what Lebi was: a game to play, not a betting product. Every later call answered to that, which kept the team away from the sportsbook mental model.
- 02
A design system from day one
I built a modular component system before the product, not after, so a small team could ship continuously. The same library later built the marketing landing pages, including a pre-launch NBA challenge, so marketing promised exactly what the product delivered.
- 03
Validate mechanics before building
I used AI prototypes to simulate the game mechanics and pressure-test them before engineering committed. Learning that a mechanic does not land is cheaper in a prototype than in a sprint.
- 04
The pivot: split the journeys
Research across four personas mapped the friction in the first prediction. Our first bet was one landing for everyone, and more than 20 interviews showed newcomers drowning in data meant for power users. I restructured the architecture into a guest discovery journey and a logged-in dashboard.
- 05
Prizes people can trust
Real prizes need real identities. Winnings stay locked until the player verifies in five steps (details, document, a selfie with the document, an email code and a confirmation), then a review; approved players cash out as gift cards on a partner platform. The mascot accompanies the key moments, so security never feels cold.
- 06
One template, any sponsor
Leagues are a template: any sponsor, any sport, any prize. Brands such as Betnacional launch their own sponsored leagues without new design work.
- 07
Handoff as a product
I delivered behavior specs, motion guidelines and edge cases, not just redlines. Treating handoff as a deliverable cut engineering's design questions by 40% and kept sprints moving.
Impact & Outcomes
- Took Lebi from scattered ideas with no identity to a shipped SaaS with a clear position: a game, not a bet.
- After more than 20 interviews, newcomers get a lighter discovery journey and power users keep their full dashboard.
- Engineering's design questions dropped by 40% once handoff became a deliverable.
- Brands launch sponsored leagues on one template: any sponsor, any sport, any prize.
A glimpse
Next case · Leaderboard
The challenge that got me hired.