I
Instrument first
Nothing ships without logs, metrics and alerting. If you can't see it, you can't trust it.
SOFTWARE · AI / ML · ROBOTICS & ELECTRONICS
Systems that don’tbreak in production.
Ganza Lab is an engineering studio in Kigali building mission-critical software, AI and automation for African industry — where most systems in production are still unreliable, manual, and quietly expensive.
GANZA LAB
KIGALI · RWANDA
SHEET 01 — CAPABILITY INDEX
Four disciplines, one accountable team.
We scope, build and operate across all four — one team, one line of accountability.
| NO. | DISCIPLINE | WHAT WE ACTUALLY DO | TYPICAL DELIVERABLE |
|---|---|---|---|
| 01 | Platform & backend | Backend, distributed systems and platform engineering. | A production system, documented |
| 02 | Applied AI / ML | Models in production, not notebooks. | A deployed model, monitored |
| 03 | Data engineering | Pipelines, warehousing and reporting you can rely on. | A governed data layer |
| 04 | Embedded & robotics | Firmware, control and physical automation. | Working hardware under software control |
SCROLL FOR REMAINING COLUMNS
SHEET 02 — HOW WE BUILD
Reliability is a design decision, not a phase.
Most systems here fail for the same four reasons. We engineer against each one deliberately, and we prove it before handover.
II
Design for failure
Power cuts, dropped networks, bad data. We assume the worst case is the normal case.
III
Automate the manual
Every repeated human step is a future outage. We replace process with software.
IV
Hand over cleanly
Documented, tested, and owned by your team — not locked to ours.
SHEET 03 — TRACK RECORD / PRIOR ENGAGEMENTS
We cut our teeth on America’s automotive backbone.
DETAILS UNDER NDA
ENGAGEMENT CONTEXT
01International teams, production scale
02Multi-year engagements
03Data engineering, ML and automation
04Client names withheld under NDA
Before Ganza Lab, our engineers spent years inside international teams — most deeply in the US automotive industry, on the systems behind servicing and warranty operations. We built the pipelines that replaced spreadsheets, the automation that replaced clerical work, and the models that told operators what was about to fail.
| DOMAIN | WHAT WE BUILT | WHAT IT CHANGED |
|---|---|---|
| Automotive servicing | Dealer, warranty and parts systems consolidated into one source of truth | One set of numbers to work from |
| RO & warranty claims | Automated validation, routing and adjudication of repair-order claims | Manual review largely removed |
| Service operations | Workflow automation across intake, scheduling and parts movement | Staff moved off data entry |
| Forecasting & reporting | Governed warehouse with demand and failure-rate models on top | Decisions the business could defend |
| Legacy integration | Bridges into decades-old OEM and dealer platforms | New capability without a rewrite |
SCROLL FOR REMAINING COLUMNS
SHEET 04 — OUR PROJECTS
The commerce layer African retail is missing.
We are building our own product. Retail and e-commerce across the continent still run on fragmented inventory, informal logistics and payment rails that barely speak to each other — precisely the class of problem the last decade prepared us for. It is in active development now, engineered to the same standard as the work we do for clients, and it is a large part of why this studio exists.
In build now, in Kigali.
SHEET 05 — WHO WE ARE
Somebody has to build it. We’re glad it’s us.
Ganza Lab is a small, senior team in Kigali. We spent years building mission-critical systems for other people's markets, and we came back with a simple conviction: the same standard belongs here. Africa's next generation of infrastructure has to be built by engineers who understand the constraints on the ground — and who actually want the job. We do.
Deep, not broad
We take a small number of engagements and go all the way in on each.
Senior hands on keys
The people who scope your project are the people who write the code.
Owned outcomes
We measure ourselves on whether the system holds up six months after launch.
Have a system that has to work?
Tell us the constraint you're up against. We'll come back with a scoped proposal, a timeline, and what we'd measure.