"The gap between a model that works in a notebook and a system that works in production is where most ML engineers stop. That gap is where I operate."
I am a Machine Learning Engineer, AI System Designer, and Data Scientist — rigorous, systematic, and wired differently. I design and build systems that sit at the intersection of mathematical depth and real-world consequence: multi-LLM orchestration pipelines, computer vision systems, NLP architectures, cloud-native deployments, and end-to-end data engineering at production scale.
I don't build prototypes. I build systems that run on real infrastructure, process tens of thousands of records, and deliver measurable outcomes in environments where failure is not an option.
That foundation is shaped by an uncommon academic combination: a Master's in Engineering AI from Carnegie Mellon University Africa, built on an undergraduate degree in Applied Physics and Computer Science. Physics gave me the discipline to model complex systems from first principles. Computer Science gave me the tools to implement them. Carnegie Mellon gave me the rigour to engineer them at scale. Together, they make me the kind of engineer who does not just apply ML — I design the systems around it.
I have mentored 100+ students across AI/ML, Data Science, and Data Analytics programmes at Moringa School, translating deep technical complexity into structured, actionable understanding for learners at every level of expertise.
|
Computational research combining machine learning with Quantum ESPRESSO simulations to predict mineral properties from first-principles physics data. Bridges computational physics and applied ML — manuscript in preparation.
|
Deep learning system for real-time crop-weed classification using CNNs and transfer learning. Designed to reduce chemical application through targeted identification, with direct application to sustainable farming at scale.
|
|
End-to-end computer vision pipeline implementing SIFT and ORB feature matching with RANSAC-based outlier rejection, and multi-band Laplacian/Gaussian pyramid blending for seamless panoramic reconstruction.
|
Sequence-to-sequence RNN with attention mechanisms for multilingual audio transcription in Swahili — a low-resource African language. Demonstrates capability in building NLP systems beyond English-dominant benchmarks.
|
|
Distributed NoSQL architecture using Cassandra for real-time data storage and retrieval, containerized with Docker and orchestrated via Kubernetes on Azure with automated scaling and load balancing.
|
Remote sensing and satellite imagery analysis for sustainable land management using Google Earth Engine. Applied spatial analysis techniques to support evidence-based environmental monitoring.
|
| Project | Description | Status |
|---|---|---|
| Basic_to_ML | A systematic series rebuilding ML and deep learning concepts from scratch — each concept first implemented at its simplest form, then extended into a full working system. Inspired by the 100-day challenge format but taken further: Snake Game ✓, Pong (next), and expanding. Not a challenge log — a structured engineering reference. | Active |
| Wolframite Classification | ML-driven mineral property prediction from quantum simulation data | Manuscript in preparation |
| AI for Science Research | Exploring AI applications in computational physics and materials science | Active |
Technical Mentor — Moringa School (Jul 2024 – Present) Designed and delivered end-to-end AI/ML curriculum covering supervised learning, deep learning (CNNs, RNNs, Transformers), NLP, and model deployment to cohorts of 100+ students. Built original teaching frameworks, debugging methodologies, and code examples from first principles. 95% student retention rate.
Materials Science Research Intern — CMU Africa (May – Aug 2025) Computational research using Quantum ESPRESSO to simulate and analyse material properties. Applied statistical modelling to interpret simulation outputs. Contributing to wolframite classification research — manuscript in preparation.
Teaching Instructor — Young Engineers Kenya (Feb – Jul 2024) STEM instruction for learners aged 5–17 covering Python, HTML/CSS, Scratch, and robotics involving sensor integration and control algorithms.
Technical Consultant — Institute of Energy Studies and Research (Sep – Nov 2022) IT infrastructure support, network management, and system optimisation across research operations.
- Wolframite Mineral Classification Using Machine Learning — Computational materials science integrating ML with Quantum ESPRESSO simulations. Manuscript in preparation, expected 2026.
- Multi-LLM Consensus Validation Architecture for Aviation Safety — Novel 5-model ensemble system with deterministic consensus engine for safety-critical classification. Research documentation on system design and validation methodology.
Languages
Machine Learning & AI
Data Engineering
Infrastructure
Visualization & BI
International Ministers Affairs Representative — CMU Africa (Dec 2024 – Present) Representing international students across institutional processes, providing emergency coordination and facilitating cross-cultural engagement.
Campus Director, Hult Prize Kenya — Multimedia University (Mar – Nov 2022) Led coordination of 20+ innovation teams competing on sustainable development challenges at the global stage.
If the work above interests you, I am worth a conversation.
