Agrimonitor
2025
A data platform for agricultural policy across Latin America and the Caribbean, redesigned for four kinds of users who don't want the same thing.
CONTEXT
Agrimonitor is an IDB platform that centralizes data on agricultural policy, productivity, food security and nutrition across the region. It needed new thematic areas and had a user base that had outgrown the original design. I led the redesign end to end as product designer, from research to implementation, as part of a studio team.
SCOPE
Full redesign
New areas: food security, nutrition, geospatial policy data Information architecture
UX/UI
30 interviews across 4 user profiles
Usability tests on the existing platform with 18 participants
THE PROBLEM
Users knew the data was there. They couldn't find it. The previous platform was built for one profile and served the other three badly: policy analysts wanted raw access and methodology, researchers wanted context, journalists wanted a snapshot they could share, sector specialists wanted depth and cross-indicator analysis. One navigation for four mental models.
DECISIONS
1. Three entry points instead of one. Country, theme, indicator: the three ways people actually think about the data, straight from the interviews. Nothing lives under a single "explore".
2. One experience, not four. Progressive disclosure: summary first, depth on demand. The specialist and the student use the same screens; the screen changes with the question, not with the user type.
3. An architecture that absorbs. Modular structure so new thematic areas plug in without fragmenting the experience. The platform was going to keep growing; the design had to assume it.
4. Data readability as the visual system. Charts, map layers and comparative dashboards tuned for reading numbers across devices, aligned with the IDB ecosystem so it doesn't drift visually over time.
WHAT I'D PUSH HARDER TODAY
Export and methodological transparency as first-class actions, not settings. The users who sign off on a number are the ones who need to trace where it came from. It's the same principle I now apply to tools for architects.