Who we are
Mizan · the balance of your data
An Ivorian data governance and valuation firm · Abidjan, Côte d'Ivoire.
Mizan (ميزان, "balance" in Arabic) is an Ivorian data governance and valuation firm, founded and led by Abdoul Aziz Touré, a Data Scientist trained at ENSAI, working between Paris and Abidjan. The name embodies three pillars: the balance between technical rigor and accessibility for business users, the analytical precision that gives weight to decisions, and the ethics that guide which mandates the firm accepts.
Its thesis fits in one sentence: "You can't build advanced models on corrupted data." The firm builds its value chain accordingly: make the data reliable first, then put it to work. That same doctrine governs the three platforms developed by Mizan Analytics: Mizan Classe for learning mathematics and physics-chemistry (in pilot), Mizan Data for business data quality and Mizan Routes for delivery routes (complete prototypes). All are built to refuse producing a result on doubtful data rather than oversell a wrong answer.
The founder
Abdoul Aziz Touré
I am a data scientist, trained in statistics at ENSAI, the French national school of statistics and information analysis. My job: make data speak without making it say what it doesn't.
In Paris, I checked the quality of waste-collection data for CITEO, built a pricing model on a consulting project at AVISIA, and I now work on AI applied to recruitment at SemantikMatch, at Station F, including measuring algorithmic bias by gender, origin and age.
I founded Mizan Analytics to bring that rigour to Côte d'Ivoire: tools that check data before using it, and advice that says honestly what the data can, and cannot, support.
Background
| 2025 → | Data Scientist, SemantikMatch (Station F, Paris): algorithmic fairness, predictive scoring, AI applied to recruitment |
| 2025 | CACAO-SECURE: forecasting Ivorian cocoa production from history, climate and prices |
| 2024 | Statistical analyst, CITEO (Paris): sorting quality, automated checks of collection data |
| 2023 | Data consultant, AVISIA (Paris): predictive pricing model and SAS dashboards |
| 2022 | Operations, CÉMOI Group (Abidjan): study of the chocolate manufacturing process and cocoa roasting |
| 2021 – 2026 | ENSAI, Master in Data Science and Statistical Engineering |
| 2018 – 2021 | MP preparatory class (advanced maths and physics), Lycée Jacques-Decour (Paris) |
Expertise
Languages: French, Bambara, English.
Sectors served
Any organization that must prove the quality of its data
What sets us apart
A verifiable commitment, not a sales promise
- Deliberate selectivity in the mandates we take on: a trust advantage in a market where trust is scarce
- Data sovereignty: deployment possible on servers of your choice, including in Côte d'Ivoire
- Transparency on the maturity of each component (available / proof of concept / exploratory), with no overselling
- Strict data isolation per organization, enforced at the database level
What's next
A second layer, built on certified data
Beyond Mizan Classe, Mizan Data and Mizan Routes, we are exploring several advanced analytics leads, still at idea stage, with no client or code today:
- Predictive Lead Scoring: an initial methodological framework was drafted with a transport/logistics-sector partner
- Marketing Mix Modeling (MMM)
- Fraud detection (diploma, background, and claims verification for employers, insurers, banks)
- Predictive logistics (stock/route optimization, distinct from Mizan Routes)
We also built CACAO-SECURE, a research prototype that forecasts annual Ivorian cocoa production from history, climate and prices, developed under the name Yvoir Futur before Mizan Analytics.
"A partner who turns down a mandate rather than oversell unreliable data, that's the promise of Mizan."