An alternative credit bureau for Nigeria's informal economy

Trace is an alternative credit bureau serving the ~60% of Nigeria's workforce that operates in the informal economy — traders, gig workers, and small-business owners who transact in cash and remain invisible to traditional credit systems like FICO or CRC.

I led the machine learning workstream. The platform issues each user a Squad-powered virtual bank account; the resulting real-time transaction stream becomes the data foundation for a calibrated credit score (300–850, FICO-comparable), delivered via a shared "Score-as-a-Service" API to microfinance lenders. The same identity model supports gig workers, whose completed jobs feed back into both credit scoring and job-matching in a single unified signal.

Built end-to-end in seven days for the GTCO Hackathon 3.0. The system integrates five Squad API surfaces (Virtual Accounts, Webhooks, Transfer, Direct Debit, Payment Gateway) and runs five ML systems across the platform: credit scoring, fraud detection, cash-flow forecasting, fairness auditing, and a multilingual job-matching engine supporting Yoruba, Igbo, Hausa, and Nigerian Pidgin.