Follow Us on Google
By Benson Michael
Founders, professionals, and graduates from across the continent converged for the Co-Create Africa Tech Exhibition 2023, an international expo sponsored by Africa’s biggest telecommunications company, MTN.
The halls were abuzz with panel discussions, product launches, and various sub-sessions. Yet the afternoon’s liveliest buzz came from a different area of fintech: AI-driven consumer lending, a critical topic in a country where economic hardship drives demand for credit while traditional banks, fixated on collateral, continue to sideline Nigeria’s vast unbanked population.
Moments after the talk, our correspondent caught up with the speaker, Brain Aboze, credit-data scientist at FairMoney. “Credit should be given where credit is due, and that means reaching the unbanked,” Brain began.
Turning thin files into full profiles
Nigeria’s consumer‑credit gap remains stubbornly wide: World Bank data suggest nearly 60 million adults have little or no formal credit history. Brain contends that alternative data can “turn a thin file into a decision-ready profile in milliseconds.”
“We don’t ask borrowers for five years of pay slips; we let the data tell their story. The goal is credit where credit is due affordable, instant, and bias-aware.”
Continental expansion: Uganda and Zambia next
FairMoney, which began as a micro-lender in Lagos in 2017, now issues 10,000+ loans daily in Nigeria. The same AI decision engine, Brain revealed, will power FairMoney’s launch in Uganda and Zambia in Q4 2023 using a “copy-and-adapt” playbook.
“The core model stays intact, but the feature universe shifts to match each market’s realities.”
AI’s exponential curve and have the macro view
While exhibition booths overflowed with GPU rigs and Web3 demos, Brain warned against fetishizing technology at the expense of economics:
Macro matters: “Rising rates squeeze lending margins. If your credit model doesn’t ingest macro signals such as FX swings and inflation, you’re flying blind.”
Exponential change: He argued that “AI isn’t merely improving; it’s compounding. Nigeria regulatory sandboxes must sprint, not stroll.”
From classroom to credit lab: advice to graduates
1. Master the fundamentals:
“Linear algebra, SQL, Python, no shortcuts.”
2. Build with real data: Kaggle is fine, “but scrape CBN rates, parse NIBSS files, feel the noise.”
3. Think product, not models. “A 0.95 AUC means nothing if the UX is clunky.”
