Senior data scientist · Machine learning engineer · Nairobi, Kenya
Building AI systems that move from research into real decisions.
I design and ship machine learning products with a strong software engineering mindset, from credit risk and lending systems to forecasting, LLM assistants, NLP research, APIs, and production pipelines.
- Senior Data Scientist at I&M Bank working on AI-driven lending and credit risk
- Cloud-aware builder across Python, SQL, Power BI, Azure, AWS, Docker, and MLOps
- Strong blend of experimentation, forecasting, LLM applications, and NLP research

Building production ML systems across credit scoring, explainability, monitoring, reporting, and financial inclusion.
About
Impact-driven, research-aware, and product-minded.
My background sits at the intersection of machine learning, data science, and business problem-solving. I enjoy taking ambiguous challenges, shaping them into measurable experiments, and turning the results into products and decision systems teams can actually trust.
Across banking, startups, education, and consulting environments, I've worked on credit scoring, lending automation, customer intelligence, recommender systems, forecasting, AI assistants, and API-backed ML services. I care as much about maintainability, deployment, and stakeholder adoption as I do about model quality.
Core ML
Data Platforms
Production
Selected Work
Featured repositories that best represent product depth and engineering range.

OfficeClaw
A production-grade Microsoft Teams automation platform with a Rust policy core, TypeScript adapter, approval workflows, auditability, and Azure deployment paths.

MedMemory
A local-first medical memory system for EHR question answering, combining FastAPI, React, Docker, and privacy-conscious AI workflows for healthcare use cases.

fastwoe
A high-performance Weight of Evidence encoding and inference library built in Rust with PyO3 bindings for Python, aimed at fast and reliable credit-risk workflows.

EduTech
A gamified learning platform for logic and mathematics with progress tracking, interactive challenges, admin workflows, and a modern Next.js stack.
Case Studies
Two examples of business-facing machine learning impact.
I&M Bank
Delivered over KES 5B in disbursements with NPL below 6%.AI-driven lending, credit risk, and financial inclusion systems for personal and SME lending.
- Built credit scoring, behavioral, and early warning models for production lending flows
- Combined ETL, model deployment, monitoring, and Power BI reporting into an end-to-end workflow
- Applied model explainability and governance practices to support business adoption
Twiga Foods
Reached 90%+ forecasting accuracy and cut manual preparation by 15 hours per week.Demand forecasting and supply optimization for a fast-moving e-commerce and logistics environment.
- Developed forecasting pipelines using Prophet, ARIMA, and LSTM-style approaches
- Automated real-time analytics workflows across Python, SQL, and Power BI
- Improved inventory and delivery decision-making through retailer clustering and logistics insights
Experience
Professional experience focused on production delivery.
Leading credit risk, digital lending, and enterprise data science initiatives that support automated decisions, portfolio growth, and proactive risk management across the bank.
- Built and deployed short-term lending models that supported over KES 10B in disbursements while keeping NPL below 6%
- Developed credit scoring, behavioural scorecards, early warning systems, and CRB-enhanced lending workflows for unsecured and micro-business products
- Delivered production data science systems, Oracle-to-Python/SQL analytics, Power BI reporting, and mentorship across model governance and business adoption
Built full-stack applications, practiced agile delivery, and shipped tested products across modern web stacks.
- Built and deployed full-stack web apps with React, FastAPI, and Django
- Mentored junior developers on agile delivery and data-driven product thinking
- Integrated Dockerized ML microservices for analytics-oriented applications
Turned operational, demand, and supply-chain data into forecasting and optimization insights inside a fast-moving e-commerce business.
- Delivered demand and supply forecasting models with over 90% accuracy
- Automated data entry and built pipelines for real-time analytics
- Improved inventory and delivery performance through cross-functional data work
Delivered business intelligence and predictive solutions, including production-ready credit scoring and anomaly detection workflows.
- Automated exploratory analysis to uncover patterns and anomalies
- Built predictive models with XGBoost and related ML techniques
- Shipped a Flask + Docker credit scoring solution for real-time analytics
Education
Academic grounding and structured technical training.

MSc, Computational Intelligence
University of Nairobi

Full Stack Software Engineering
Moringa School

BSc. Actuarial Science
University of Nairobi

Student Actuary
Institute and Faculty of Actuaries
Certifications
Credentials that reinforce cloud and machine learning practice.
Claude Code in Action
Anthropic
Oracle Database 19c: Basic SQL
Natural Language Processing - Transformers
Hugging Face
AWS Machine Learning Engineer Nanodegree
Udacity
Machine Learning in Python with Scikit-Learn
INRIA - The French national research institute for digital science and technology
Google Cloud Certified Professional Cloud Architect
Microsoft Certified Azure Fundamentals
Microsoft
Amazon Web Services Cloud Practitioner
Amazon Web Services
Apache PySpark by Example
Data Engineering Pipeline Management with Apache Airflow
LinkedIn Learning
Machine Learning Engineering for Production (MLOps) Specialization
Coursera
Retrieval Augmented Generation (RAG)
DeepLearning.AI
Contact
Open to thoughtful teams, ambitious products, and meaningful ML work.
If you're building something that needs machine learning, experimentation, NLP, or a strong production mindset, let's talk.



