Four products, four stages.
The portfolio is complete. Each product holds a layer of the stack and carries its stage in its own sentence.

The portfolio is complete. Each product holds a layer of the stack and carries its stage in its own sentence.
Planche 02Pillar
In productionTelecom rails
Pillar puts airtime, data, SMS and mobile money behind a single API. In production across the UEMOA zone and the Gambia.
One integration replaces as many integrations as there are operators. Catalogue, balances, transactions and their tracking all go through the same interface, with the same request grammar.
This is the product that needs no one's permission to make its case: it runs, with its own domains and its own API.
- CoverageUEMOA zone and the Gambia
- Brings togetherAirtime, data, SMS, mobile money
- InterfaceA single API
- StageIn production
Weaver
In developmentTelecom rails
Weaver delivers your business messages across multiple operators. You can check delivery, message by message. In development, for Benin, Côte d'Ivoire, Mali and the Gambia.
Verifiability is the point of the product. What was delivered can be observed, message by message, rather than inferred from a total.
Weaver is described by its results: what is delivered, the coverage, and what you can check for yourself.
- Target coverageBenin, Côte d'Ivoire, Mali, the Gambia
- PurposeBusiness messages across operators
- VerificationDelivery observable, message by message
- StageIn development
Xonou
At design stageMoney rails
Xonou: a wallet, a marketplace, a card-to-mobile-money bridge, in a single product. At design stage.
Three uses that today live in three separate tools will be brought together in a single product.
Xonou is at design stage. What follows will be written in the present tense the day it is true.
- Will bring togetherWallet, marketplace, card bridge
- LayerMoney rails
- StageAt design stage
MaVoix
In scopingIntelligence rails
MaVoix makes electoral data queryable in natural language. The system says when it doesn't know. In scoping.
Ask a question in plain language and get an answer drawn from public data, with its source.
The sentence that matters is the second one: the system says when it doesn't know. Admitting a limit beats a confident wrong answer.
- PurposeQueryable electoral data
- InterfaceNatural language
- BehaviourThe system flags what it doesn't know
- StageIn scoping