We design and engineer digital products across AI, finance and education — from payment platforms to the apps people open every day. Built in Cameroon, for teams anywhere.
01Capabilities
Different domains, one standard of engineering. The team that ships a payment ledger also reviews the code behind a learning app.
LLM products, agents and evaluation pipelines that hold up outside the demo.
Agents · RAG · EvalsPayments, mobile money, lending and ledgers built to reconcile to the franc.
Ledgers · PaymentsLearning platforms, assessment and school operations for real classrooms.
Learning platformsNative and cross-platform apps that stay useful on a weak connection.
iOS · Android · OfflineInternal platforms, integrations and workflow systems that replace spreadsheets.
Workflow · IntegrationWeb applications and APIs engineered for speed, clarity and scale.
Next.js · APIs03Services
Six practices, one team. Most engagements draw on several at once — an AI feature is also a product, a payment flow is also a design problem.
How we work04Built at CMBYTES
Client work teaches us where the gaps are. Sometimes we fill them ourselves.
All products
Group savings, run like clockwork.
Kolo brings the njangi — the rotating savings group millions of Cameroonians already trust — onto a shared ledger. Contributions, payout order, penalties and meeting notes are reconciled against mobile money, so the treasurer's notebook becomes a record every member can see.
05Engineering
Software rarely fails at launch. It fails slowly, through shortcuts that compound. We design for a system’s third year, not its first demo.
Code a new engineer can change with confidence: clear module boundaries, typed interfaces and decisions written down.
Latency and payload budgets set up front, and measured on the phones and networks your users actually have.
Systems that grow by adding capacity rather than rewrites — without distributed complexity before it's needed.
Threat models before launch, least-privilege access, secrets out of code and audit trails wherever money moves.
Deliberate, documented trade-offs. The simplest design that meets the requirement, with a clear path beyond it.
06From CMBYTES
Engineering, AI, fintech and technology in Africa — written by our team, not a content agency.
Read the publicationThe prototype is the easy part. What separates an AI feature people trust from one they quietly stop using is everything around the model: evaluation, failure handling, and the humans in the loop.
Eyong Tabe4 min read
