AfriyonTechPlatform
About

We build software for the parts of the business nobody photographs

Invoicing runs, stock reconciles, records stay complete. AfriyonTech exists because those unglamorous processes decide whether an operator makes money — and they are usually the first thing broken by generic software.

3
Production products
6
Countries served
99.9%
Target availability
2021
First deployment
Why we exist

Generic software loses to operational reality

When a system does not match how a business actually runs, staff build a shadow process around it. The spreadsheet returns. The data splits. Decisions get made on numbers nobody trusts.

We started inside an ISP, not a software company. The first version of our billing platform existed because a reconciliation process was consuming three people for a week every month, and the numbers it produced were still disputed.

That constraint shaped everything since: build the system an operator will actually use at 7am on a Monday, prove it against their own baseline, and be honest about where it stops.

  • Products are deployed independently, not as one monolith
  • Every customer gets an isolated tenant, by design
  • Healthcare workloads run on dedicated databases
  • Demo environments never touch production data

How we work

  1. 2021

    Built for one ISP

    The billing platform started as an internal tool for a 3,000-subscriber operator that could not reconcile mobile-money payments reliably.

  2. 2023

    Second product, different industry

    Pharmacy groups asked for the same discipline applied to stock and expiry. The platform layer — identity, tenancy, billing — became shared.

  3. 2024

    Clinical workloads

    AfriyonCare HMIS entered production with a dedicated-database architecture and HL7 interfaces to laboratory systems.

  4. 2026

    One platform, three systems

    A unified customer portal and demo centre now front all three products, with independently deployable release cycles.

Principles

The decisions we keep making the same way

These are not values on a wall. They are the trade-offs we default to when a project gets difficult.

01

Depth over breadth

We would rather own three workflows completely than touch thirty superficially. Each product is judged by whether an operator can stop keeping a spreadsheet beside it.

02

Boring reliability wins

Novel architecture is a cost the customer carries. We choose the well-understood database, the boring queue and the tested backup, then spend the innovation budget on the workflow itself.

03

Say what the product does not do

Overpromising is a debt paid by the support desk and the customer. Our proposals include the limits of the system in writing, before the contract.

04

Isolation is not a feature toggle

Tenant separation is enforced at every layer or it does not exist. Healthcare customers get a dedicated database because a shared schema is the wrong trade for that data.

Security posture

Claims we can support with a document

We will not claim certification we have not earned. What we can do is describe exactly how the platform is built and let your assessor verify it.

TLS everywhere

HSTS enabled, secure headers by default.

RBAC + MFA ready

Per-product roles, optional TOTP second factor.

Encryption at rest

Volume-level encryption on managed databases.

Audit trails

Administrative and clinical access is attributable.

Rate limiting

Applied per route class, not just globally.

Tested restores

Backups are restored on a schedule, not assumed.

Note on compliance: we design the platform to support applicable data-protection and healthcare regulations in each deployment jurisdiction. We do not claim certification or accreditation for any product without independent assessment in that jurisdiction.

Where we work

Nairobi-based, deployed wherever the customer runs

The team operates from Nairobi with deployments across East and Southern Africa. Support coverage follows the deployment region.

Head office
Nairobi, Kenya
Engineering
Nairobi · remote
Support hours
Mon–Fri, 08:00–18:00 EAT
Critical incidents
24/7 on enterprise plans

Talk to the people who build it

No account managers reading a script. A working session with an engineer who can answer questions about your actual workflow.

Synthetic demo data only. No production credentials are ever exposed.