I’m a Machine Learning Engineer and Mathematician passionate about using intelligent systems, graph theory, and optimization to solve real-world African problems.
With a background in pure mathematics and applied data science, I focus on building models and tools that make infrastructure smarter, healthcare fairer, and development more data-driven — all grounded in local context.
I believe machine learning should be more than algorithms and benchmarks — it should solve problems that actually matter to people. That’s why I build systems that are:
- Context-aware: grounded in the realities of African cities, households, and public institutions
- Theory-informed: backed by graph theory, linear algebra, and optimization
- Practical: built with code, not just papers — and meant for real users
- Impact-oriented: focused on equity, accessibility, and better decision-making
I enjoy bridging the gap between mathematics, machine learning, and social impact — especially in public health, transportation, and infrastructure.
Languages & Libraries
Python, NumPy, Pandas, scikit-learn, NetworkX, Matplotlib, Seaborn, R, CVXPY
Core Skills
Graph Algorithms, Supervised Learning, Unsupervised Learning, Linear Programming, Data Wrangling, Model Deployment (basic)
Other Tools
Jupyter, Git, Excel, Power BI, Markdown, LaTeX
| 🚀 Project | 🔍 What It’s About |
|---|---|
| Functional Digraph Decomposition (Undergraduate Research) | An academic exploration into decomposing functional digraphs into rooted trees. Focused on theoretical graph concepts with detailed proofs and algorithm analysis. |
| Ant Social Colonies | Exploring How Ant colonies can be modelled using Graph Theory |
| NatView Chatbot | Developed a chatbot for NatView Foundation to support user interactions and automate responses. Implemented using [Python, LangChain, Streamlit, OpenAI API], focusing on natural language processing and user experience. |
| Graph Theory for AI Security: Detecting Prompt Injection | Applying graph theory techniques to detect and prevent prompt injection attacks in AI agents — enhancing robustness and security in conversational AI systems. This project is currently in progress. |
- Building more ML pipelines for development-focused use cases
- Using Graph Theory to understand Attacks on AI Agents
- Publishing more on graph-based ML and African data equity
“Data isn’t neutral — but used wisely, it can move us closer to equity, insight, and impact.”
I’m always up for meaningful collaborations — especially if you’re working at the intersection of:
- Machine learning and the public sector
- Infrastructure data and algorithmic design
- African tech ecosystems and applied research
Reach me at: nanfeyarnap3@gmail.com
Or find me on: LinkedIn
Thanks for stopping by. This portfolio is growing — just like me.