Small perturbations, cascading consequences: that's the core lesson of chaos theory, and exactly why interpretability matters when we hand decisions to models we don't fully understand.
I study Mathematical Engineering at ENSGMM (Benin), focusing on the intersection of applied math, dynamical systems, and deep learning. My main interest is AI safety and mechanistic interpretability : understanding why models behave the way they do, not just confirming that they work.
I'm a researcher with an entrepreneurial background. Before fully committing to math, I spent a year building and running a marketing agency (Kreators Funnels). I know how to ship, market, and sell products, not just theorize about them.
- Currently: Modeling dynamical systems and exploring deep learning.
- Writing: I publish essays on AGI risk, algorithmic control, and cognitive agency on my Science Blog, and Substack.
- Interests: Sci-fi, psychology, and the control problem.
Languages:
Data & Modeling:
Design & Growth:
| Project | Stack / Methods | Focus & Summary |
|---|---|---|
| The Lorenz System | Julia, Runge-Kutta |
Deep study of 4D chaotic dynamical systems and cryptographically secure PRNGs. Features 3D strange attractor visualizations and Lyapunov sensitivity analysis. |
| SEIRS Spatial Heterogeneity | Python, JupyterLab |
Research report from the NSRI Summer 2026 Hackathon. Modeling SEIRS epidemic dynamics considering spatial heterogeneity and population connectivity. |
| Regression Cox Model | R, Biostatistics |
Statistical study and implementation of Cox proportional hazards survival model on the 'cancer' dataset with Schoenfeld residual diagnostics. |
| Poisson Regression (GLM) | R, GLM |
Comprehensive Poisson regression model on the Airquality dataset. Overdispersion testing, goodness-of-fit diagnostics, and rigorous interpretations. |
| SciNexus Backend | Java, Spring Boot |
Backend architecture for the SciNexus scientific blogging platform. Modular domain modeling, data persistence, and clean RESTful API design. |
| Four in a Row | Java, JavaFX |
Complete OOP desktop implementation with modular MVC architecture. |
- "AI, Influence and the Future of Humanity" : On AI's societal trajectory, cognitive dependence, and the control problem. (Feb 2026)
- "AI: Towards the Dehumanization of Humanity?" : Examines macro-societal AGI risk and the erosion of human agency. (Dec 2025)