Nairobi, Kenya · 2024
Sevi: explainable credit for Kenyan retailers
ML credit decisions, the governance around them, and the full app design.
Sevi is a CBK-regulated fintech that finances stock purchases for retailers across Kenya. I built its AI credit verification system from zero: the ML pipeline, Kubernetes infrastructure, API and decision dashboard. I also designed the full customer app.
Around the models sits a governance layer of explainability, bias monitoring and audit trails, so credit decisions in thin-data markets are fast, fair and defensible. Today, as Chief Data Officer, I lead the data department and own how data becomes product and credit decisions.












