Skip to content
D
All work

EdTech · Flutter · Firebase · Product

Making academic resources accessible to 25,000 students.

Co-founder and product & technical lead of an academic support app for Cameroonian students: a bank of past exams, an AI tutor, quizzes, a peer network and a payment flow tailored to the local market.

Role
Co-founder · Product & technical lead
Period
2021 — present
Stack
FlutterGetXFirebasePythonTensorFlowLLM API

Measured reference points

25,000+

registered users

EdTech product co-founded and led technically

≈ 1,000

daily active users during exam periods

Peak measured in 2024–2025 sessions

10,000+

learning resources

Past exams, solutions and vetted notes

19

releases shipped

On iOS and Android since 2023

The problem

In many Cameroonian universities, past exam papers are essential to prepare for the finals. They circulate as photocopies, remain hard to find and are unevenly available depending on the school or the city. Students also have very few options to get quick help when they get stuck on a problem.

Product ambition

Put in every student’s pocket a base of academic resources, a peer help space and an assistant available at any time — without mechanically copying EdTech models designed for other markets.

My contribution

Product — feature definition and evolution, roadmap trade-offs, adaptation to real usage, integration of AI in user flows, tracking of adoption metrics.

Mobile — Flutter architecture, state management and navigation with GetX, Firebase integration, notifications, deep links, caching and optimization for low-end Android devices, store publishing and updates.

Leadership — coordination with my co-founder and the team, distribution of technical topics, architecture decisions, coherence between product, data and mobile experience.

Key decisions

Designing for real networks and devices

The app needed to work on low-end Android phones and in areas with unstable connectivity. Caching, resource sizing and local persistence were treated as product decisions, not secondary optimisations.

Adapting monetisation

Card payments do not match the habits of most targeted students. Purchase flows were designed around Mobile Money, with fallbacks so that a user is never blocked for purely technical reasons.

Keeping AI server-side

Model calls are orchestrated server-side to protect keys, control costs, cache some responses and allow us to switch providers without shipping a new mobile version.

Results

VALIDE has passed 25,000 registered users and reaches around 1,000 daily active users during exam periods. The product received the Cameroon AI Impact Award 2026 and has built partnerships with several university organisations and communities.

What this project proves

  • carrying a product over several years
  • Flutter and Firebase at real scale
  • understanding a local market
  • balancing impact, cost and complexity
  • product vision combined with technical responsibility

A similar need on your team?

Let's talk about the context, the product and how I can contribute.