Mission
To build accessible, locally-grounded technology that solves real problems for African communities — and to train the engineers who will keep building it.
VAF UBWENGE TECH
The Company
How a group of data science students in Kigali turned a classroom idea into a working technology company.
2023
Founded in Kigali
12+
Students trained
6
Products shipped
100%
Built in Rwanda
Our Story
VAF UBWENGE TECH began in a lecture hall at Université Libre de Kigali. A handful of data science students kept running into the same frustration: the tools being taught were built for problems somewhere else. Transport, payments, language, logistics — the systems people actually use every day in Rwanda were either missing or borrowed from contexts that did not fit.
So the group started building instead of waiting. The first prototypes were rough — a bus-tracking board, a scoring script, a Kinyarwanda chat experiment — but they worked, and people used them. That was the proof the team needed.
Today VAF UBWENGE TECH operates as a working studio: an intelligence lab that researches and trains models, and a product team that ships them. Everything we release is built, tested and maintained by students and recent graduates who live with the problems they are solving.
Vision
A generation of Rwandan engineers who do not have to leave to work on hard problems, because the hard problems — and the companies solving them — are here.
How We Work
We start from the field, not the framework. Every project begins with time spent watching how something is actually done — a bus queue, a market stall, a clinic intake desk — before a line of code is written.
We build small and ship early. A rough version in real hands beats a polished version in a slide deck, and it tells us within a week whether the idea holds.
We train while we build. Every project carries at least one student who has never shipped before, because the point is not just the product — it is the engineer who comes out the other side.
What We Stand For
01
Built for here
We design for Rwandan realities first — language, connectivity, payment rails and cost — rather than adapting something built elsewhere.
02
Learning is the product
Every project trains someone. The code ships, and so does the engineer who wrote it.
03
Show the work
Open demos, public leaderboards and honest write-ups. If it only works in a slide, it does not work.
04
Useful beats impressive
We would rather ship a plain tool people rely on than a clever one they try once.
What We Do
Applied machine learning
Model training, evaluation and deployment — from Kinyarwanda language models to computer-vision classifiers that run in the browser.
Product engineering
Full-stack web and mobile products, built to work on the connections and devices people actually have.
Data systems
Pipelines, dashboards and scoring engines that turn raw operational data into something a team can act on.
Research & training
An open intelligence lab, public competitions and a leaderboard where students learn by shipping real models.
Milestones
2023
The first prototype
A bus-tracking board built for a class project turns into BusTag.
2024
The Intelligence Lab opens
In-browser model training and inference, free for any student to use.
2024
WANDAA AI
Work begins on a Kinyarwanda-first conversational model.
2025
Competitions go live
A public leaderboard and scoring engine for student ML challenges.
A generation of Rwandan engineers who do not have to leave to work on hard problems, because the hard problems — and the companies solving them — are here.