Building production-ready machine learning systems from data to deployment.
I'm a Machine Learning Engineer and Data Scientist focused on building reliable, production-oriented ML systems.
My work spans the complete ML lifecycle — from data preparation and experimentation to model deployment, orchestration, monitoring, and CI/CD.
- Build and deploy production-grade machine learning systems
- Develop REST APIs for ML inference using FastAPI
- Design end-to-end MLOps pipelines
- Track experiments and models with MLflow
- Version datasets and ML artifacts using DVC & DagsHub
- Automate workflows with Prefect
- Containerize applications using Docker
- Implement CI/CD pipelines with GitHub Actions
- Work with cloud and infrastructure technologies including AWS, Kubernetes & Terraform
- Develop data-driven solutions using Python, SQL, Pandas & PySpark
Production-ready machine learning system for real-time transaction risk scoring.
Performance
- 79% F1-score
- 81% precision
Engineering
- XGBoost + SMOTE for imbalanced classification
- DVC for data versioning
- Prefect for workflow orchestration
- MLflow for experiment tracking
- FastAPI for model serving
- Docker for containerization
- GitHub Actions for CI/CD
- Deployed on Render
Focus: Production ML · MLOps · Model Serving · CI/CD
Containerized ML API designed to turn customer predictions into actionable retention signals.
Performance
- 84% accuracy
- 63.1% weighted F1-score
Engineering
- LightGBM
- DVC + DagsHub
- MLflow
- Prefect
- FastAPI
- Docker
Focus: Customer Analytics · MLOps · API Deployment · Experiment Tracking
Data-driven analysis of passenger lounge demand, eligibility and revenue opportunities.
Highlights
- Analysed 50,000+ flight records
- Modelled passenger eligibility across 3 lounge tiers
- Identified peak demand of approximately 1,280 users/hour
- Built a revenue model projecting £262.8M annual profit
- Recommended Concorde Room expansion
- Estimated 2-month payback on a £7–10M investment
Tools: Python · Pandas · Seaborn · Power BI
Focus: Data Science · Business Analytics · Revenue Modelling
Machine learning classification pipeline for predicting flight booking completion.
Highlights
- Analysed 10,000 customer records
- Achieved 61% accuracy over baseline
- Engineered 8 domain-specific features
- Built features around booking lead time, group size, long-haul travel and extras
- Developed recommendations targeting improved conversion and reduced marketing costs
Tools: Python · Scikit-Learn · Random Forest
Focus: Predictive Analytics · Feature Engineering · Business Intelligence
Eastern Africa Statistical Training Centre (EASTC)
B.Sc. Data Science Dar es Salaam, Tanzania
Core Areas
Data Science · Machine Learning · Statistical Modelling · Data Engineering · MLOps · Model Deployment · Production AI
I'm interested in:
- 🤖 Machine Learning & AI opportunities
- ⚙️ MLOps & ML Engineering roles
- ☁️ Cloud-based ML systems
- 🤝 Freelance projects & collaborations
- 🌍 Remote opportunities
Building reliable ML systems that move from experimentation to production.