A ton of content. No world to hold it.
By 2020, Escola da Inteligência — a company created by leading names in socio-emotional education in Brazil — had accumulated a deep content library: drawn characters sorted by age, by school grade, and by emotion, tied to years of classroom lessons. The brief was to turn that library into a game with LMS features, with every illustration produced in-house by an illustrator I led.
The hard part was the intersection. Each grade came with its own characters, its own lessons, its own emotions, and its own OKRs — and the players ranged from kids to teenagers, whose research needs differ by persona and age range. On top of the design work, the project meant managing the team and reporting directly to the CEO.
Infrastructure first, worldbuilding second
Two decisions defined this project, and neither of them looks like game design.
First: before designing a single screen, I built an atomic research structure — a system able to sustain every need for end-user input, deliberately scoped to serve all upcoming projects, not just this one. Research here was not a phase on a timeline; it was infrastructure the whole company could draw from.
Second: I wrote a game statement — one scope document fixing platforms, target audience, game objective, monetization models, benchmark, pedagogic context, player experience and more. Every effort that followed pointed back at that document.
Research was built as infrastructure, not scheduled as a phase.
Building the city, block by block
-
One game per grade, mapped as an intersection
I started by mapping a game for each school grade — understanding where its characters, lessons, emotions and OKRs intersected — and mapped the research needs per persona and age range across kids and teenagers. Then I amassed the available content, starting with the actual characters tied to the grade prioritized for the game.
The grade-by-content intersection map — the MVP slice highlighted. -
A game statement as the contract
The scope document fixed platforms, target audience, game objective, monetization models, benchmark, pedagogic context and player experience in one place — written explicitly to direct all efforts. When a company has this much content, the risk is not running out of ideas; it is drowning in them.
-
Patterns for a 4th-grade player
For the first MVP — 4th-grade kids — I created the design patterns: style sheets, component sheets, everything UI and front-end development would need to build consistently. The same discipline I bring to enterprise design systems, aimed at 4th-graders.
-
The corrosion mechanic
I created patterns for in-game corrosion: certain actions from the player — or inaction — would degrade the game's structures. The world does not wait for the player; it responds to being neglected.
The corrosion pattern — action, or the lack of it, changes the state of the world. -
Mapa Mundi, neighborhoods, and the shape of a path
The game map splits into two layers: the Mapa Mundi, giving access to the neighborhoods, and the neighborhoods themselves — each one representing an emotion the kids encounter in class. For every neighborhood I devised a design pattern instilling a feeling of choice through different paths, rewarding curiosity with easter eggs, and dividing checkpoints to distribute cognitive load equally across the path.
Two layers of the world — Mapa Mundi for access, neighborhoods for one emotion each.