AI
Voice assistant in national languages
A speech recognition and synthesis prototype in Mooré and Dioula, led by the Data & AI Lab.
Our five principles
- 01
Useful
AI is justified by a real problem and a measurable gain, not by hype.
- 02
Reliable
No AI without quality data: that's always where we start.
- 03
Frugal
The simplest model that works, at the lowest cost, with the smallest footprint.
- 04
Sovereign
Your data stays under your control, hosted to your requirements.
- 05
Human
AI assists, humans decide: supervision, explainability and the right to appeal.
In-house
Consultants augmented by agents
Our teams work with agents that make data reliable (anomaly detection, quality checks, automated reports), prepare analyses and document projects. The time saved is reinvested in advice and client relationships.
- An AI usage charter signed by every employee
- No client data in an unapproved tool
- Prompt Engineering training for all teams
AI
AI agent for data regulation monitoring
An agent that tracks data-protection texts in West Africa and answers compliance questions.
Insight
Article
Article
Insight
White paper
Infographic
Podcast
An AI use case in mind?
Let's assess its feasibility and value together.