The Decolagem project, run by Gerando Falcões, supports a network of NGOs across Brazil that work directly with families in underserved communities. The tool at the center: a mobile app used by mentors and volunteers to run community diagnoses — mapping families' needs so NGOs can allocate resources where they matter most.
The problem was scale and friction. The original diagnostic flow had more than 180 questions and took over an hour per family. For mentors working in the field — often in areas with limited connectivity and high emotional intensity — that length was unsustainable. Families dropped off. Data quality suffered.
What we did
We started by interviewing mentors and volunteers to understand where the existing flow broke down. The most common complaint: many questions felt redundant or irrelevant to specific community contexts. The diagnostic tried to cover everything at once, for every case.
The redesign focused on three moves:
Prototypes were tested in the field — not in a lab — with real mentors in real communities. Feedback drove several rounds of iteration before the final version.
Results
The constraint that shaped the project: users weren't tech-savvy, the environment was unpredictable, and the stakes were real. That forced every design decision to prioritize clarity over cleverness.