Develop practical capabilities in scoring thin-file and no-history borrowers.
Turn mobile money, airtime and other alternative data into a credit scorecard for thin-file and no-history borrowers. You engineer features without data leakage, test stability, and prepare documentation for regulators and investors.
Practical modelling with Python and AI-assisted development
Use AI tools, including Claude Code, to accelerate model development, testing and documentation.
Understand scoring for thin-file and no-history borrowers
Structure mobile money, airtime and behavioural data
Engineer features while preventing data leakage
Build a scorecard for thin-file borrowers
Test whether the scorecard performs as expected
Document the scorecard for external review
Use AI to speed up the build.
Use testing and reconciliation to prove the numbers.
Use MFI credit expertise to make the judgement.
You will have worked through a practical end-to-end AI-assisted MFI scorecard build, from mobile money and airtime data through feature engineering, calibration, validation and reporting for regulators and investors.
| Duration | 2 days |
|---|---|
| Delivery | Instructor-led with live demonstrations and practical exercises |
| Investment | US$200 per delegate |
| Group bookings | 20% discount |
| Prerequisites | Laptop, working knowledge of credit risk and Excel |
| Tools | Python and AI-assisted development, including Claude Code |
| Facilitator | Tafirei Mashamba, PhD |
| Next intake | Ask us for the next intake. We reply by email or WhatsApp. |
Need this built for your institution instead? See Credit Risk Modelling.
Book your seat, or ask about training your whole team. Group bookings receive a 20% discount.