I build production-ready data systems for monitoring, evaluation, accountability, and learning (MEAL) in complex operational environments. My focus is on real-world systems that help UN agencies, NGOs, and East African operators turn fragmented data into decisions, reporting, and measurable impact.

• M&E and humanitarian data systems for field operations, programme monitoring, and donor reporting.
• Applied AI tools for risk analysis, profiling, insights, and decision support in high-context environments.
• Data pipelines and reporting systems that work in environments where quality, traceability, and governance matter.
• Business and operational platforms for East African institutions, including POS, ERP, and offline-first workflows.

• gcf-food-security-intelligence — Climate-informed food insecurity risk prediction and early-action intelligence.
• un-review-hub — Federated review workflows with data sovereignty and human-in-the-loop governance.
• un-sdg-east-africa — SDG monitoring and analytics for East African policy and programme decisions.
• responsible-dq-engine — Survey data quality and privacy-risk auditing for ODK/KoboToolbox ecosystems.
• kenya-malnutrition-risk-prediction — Predictive analytics for acute childhood malnutrition risk in Kenya.

• sdg-rag-system — Retrieval-augmented generation grounded in SDG and urban transport frameworks.
• intelligent-profiling-engine — Hybrid AI profiling combining unsupervised learning, NLP, and human review.
• humanitarian-data-project — Reproducible analysis of public international datasets for development and humanitarian use cases.

• mbumah-hardware-pos — Multi-tenant POS and ERP for Kenyan retail operations.
• retail-erp-pos-kenya — Offline-first retail operations with M-Pesa and eTIMS considerations.
• dukaflow-mobile — Flutter-based POS experience for Android and iOS.
• hardware-pos-erp — Inventory and operational workflows for hardware retail.

• python-raft-consensus — From-scratch implementation of the Raft consensus algorithm.
• blockchain-inventory-system — Inventory management on an immutable ledger with AI-assisted workflows.
• data-engineering-analytics-platform — Analytics platform built with modern frontend and backend stacks.

Explore all public repositories →

1. Define the problem precisely before building.
2. Design around field realities, governance, and reporting requirements.
3. Deliver in stages with measurable milestones and clear checkpoints.
4. Document and hand over systems that teams can trust in production.

This approach keeps technical work grounded in operational realities rather than abstract demos.

For M&E data systems, evaluation-support automation, donor reporting, or humanitarian data engineering, connect via the portfolio or X / Twitter. You can also open an issue on any public repository here and it will reach me.

_{Profile index and selected public repositories reflect a portfolio shaped around data systems for impact, governance, and operational delivery.}