V
Loading...
Loading...
Loading...
Financial Access Through AI
ML models using mobile data and transaction patterns enable loans for millions who lack traditional credit history, driving financial inclusion across Africa.
3x
More Approvals
15M+
Users Enabled
<5%
Default Rate
Equal
Approval Rates
Over 60% of African adults lack formal credit history, making them invisible to traditional credit scoring systems. Without credit, they can't access loans to grow businesses, handle emergencies, or invest in education.
We developed alternative credit scoring models that use mobile phone usage patterns, mobile money transaction history, and other alternative data sources to assess creditworthiness. Our models enable lenders to make informed decisions about borrowers who would otherwise be rejected.
The result: millions of previously unbanked Africans can now access credit, fueling small business growth and economic development across the continent.
Understanding the unique obstacles we're working to overcome.
Traditional credit bureaus have no data on the majority of African adults.
Many potential borrowers work in the informal economy with irregular income patterns.
Using alternative data raises privacy concerns that must be carefully managed.
Models must not discriminate against vulnerable groups while maintaining predictive power.
The methods and techniques we've developed to address these challenges.
Mobile money transactions, airtime purchases, and phone usage patterns as credit signals.
Models trained with fairness constraints to ensure equitable access across groups.
Clear explanations for credit decisions to build trust with users and regulators.
Ongoing monitoring for model drift and fairness violations in production.
Measurable outcomes from our research and deployments.
3x
More Approvals
Alternative scoring enables 3x more loan approvals vs. traditional methods.
15M+
Users Enabled
Over 15 million previously unbanked users have accessed credit through our models.
<5%
Default Rate
Loans enabled by our scoring have default rates under 5%, similar to traditional lending.
Equal
Approval Rates
Model achieves equal approval rates across gender and urban/rural demographics.
Mwangi, J., Okonkwo, A., et al.
Expanding alternative credit scoring to 10 African countries with telco partners.
Partners:
Safaricom, MTN, Airtel, Orange
James Mwangi
Technical Lead
Dr. Amara Okonkwo
Fairness Advisor
We're always looking for collaborators, partners, and talented researchers to advance this work.