NABU works to end illiteracy by publishing storybooks in kids' native languages, and they're reaching over 12 million learners worldwide. We designed the platform behind all that data, taking three disconnected systems (MySQL, spreadsheets, and Google Analytics) and consolidating them onto one centralized platform for admins to manage partners, organize data, and build visualizations that actually show NABU's impact.
SHORT ON TIME?
TIMELINE
Feb 2025 - May 2025
TOOLS
Figma
ROLE
Product Designer
SKILLS
Product Design
Data Visualization
Design Systems
Problem
NABU helps kids around the world learn to read in their native languages through fun, digital storybooks. But behind the scenes, their data was all over the place, split between MySQL, spreadsheets, and Google Analytics. Every time a new partner joined, admins had to manually set up dashboards and chase down data from country managers. It was messy, time-consuming, and hard to keep everything consistent. Our goal was to design one simple, centralized platform that made it easy to manage data, track impact, and keep NABU's mission running smoothly.
Understanding the Users
PERSONA
Leila Mwangi
ABOUT
NABU Admin
Struggles with sorting and filtering through disorganized datasets across multiple systems
“I spend more time cleaning data and setting up dashboards than actually analyzing our impact. I just want everything in one place.”
PAIN POINTS
Juggling multiple data sources (MySQL, spreadsheets, Google Analytics) is time-consuming and confusing
Creating new Looker Studio dashboards for every partner leads to redundancy and clutter
Manual data submissions often have errors or missing values
No centralized system to view, approve, or manage partner accounts and roles
NEEDS:
Leila needs a single platform that brings all of NABU’s data together, so she no longer has to juggle spreadsheets, databases, and analytics tools. She wants an easy way to approve partners, manage roles, and assign dashboards, all while keeping NABU’s branding consistent.
Creating a Design System
NABU gave us a very thorough design system and clear branding guidelines, which meant we could focus our time on solving the actual data problem instead of rebuilding the basics. Since we already had so many great resources, we only had to design a few extra components, letting us move faster toward something NABU could actually put to use.
Project Considerations
Our project deals with a lot of data, so I spent a lot of time researching how to best organize it to optimize the workflow for creating visualizations.
We were on a tight timeline, with only three months to design and put our application into production. With this in mind, I made sure to prioritize efficiency without sacrificing design quality, so NABU would have a solid foundation to build on going forward.
This project was mostly about turning NABU's data into well-organized tables, which is different from the typical UI/UX projects I usually work on. That meant a lot of learning, not just to understand the data itself, but how admins and partners actually interact with it, so I could design something that held up under how it would be used.
Solution
The NABU dashboard we built is a centralized web platform that lets admins manage partner access, organize data from multiple sources, and build custom visualizations, giving NABU one reliable place to make decisions and support their partners.
Review Partner Access Table
NABU wanted a platform that better organized and centralized their data, so one of the solutions we proposed was using tables. We designed the Review Partner Access table, which lets NABU administrators approve, reject, and view details on each requestor.
Multi-select checkbox lets admins approve or reject requestors in bulk instead of one at a time
Search bar helps admins quickly find a specific requestor instead of scrolling through the full list
Role Builder Table
This table helps organize the different roles partners and volunteers can have. Roles control access to data and visualizations, so volunteers only see what’s relevant to them, keeping sensitive data protected while still letting NABU share the right insights with the right people.
Expanding visualizations within a dropdown lets users view several at once without cluttering the table
Edit, delete, and create role buttons put all the key admin actions in one place, so managing access takes minutes instead of back-and-forth
Add Partner Flow
The pop-up modal design cuts down on the number of tabs a user needs open, and lets them see existing roles they’ve already created, so onboarding a new partner takes a few minutes instead of turning into its own task
Visualization Builder
This breaks down the process of turning data into a visualization step by step, simplifying the process for admins who lack technical experience.
Splits the process into 5 steps, keeping it organized and easy to follow
Pre-filled dropdown menus simplify the process and cut down on room for error
Data Visualization Examples
The Visualization Builder was one of the most important parts of our design, since it’s what lets admins turn scattered data into something partners and funders can actually read and trust. Here are a few examples of what that data could look like once it’s organized.
Takeaways
Designing for a data-heavy, internal application
This was my first time designing an internal application built around this much data, and it changed how I think about data organization and just how much of an impact it can have on workflow efficiency at scale.
Designing for internal efficiency
Every screen I designed had to make an admin’s day-to-day easier, not just look good. Things like bulk partner approval and role-based dashboard access weren’t just features, they were direct fixes to the manual, error-prone workflow NABU was stuck in. I learned to prioritize that kind of efficiency first and then focus on visual appeal.
Building technical fluency to support creativity
Understanding how the data is aggregated and filtered gave me more room to be creative, since I could focus on designing a visualization builder that let admins go from raw, scattered data to something a partner or funder could actually understand.