Fanaka Technologies Limited is a Zambia-based fintech on a mission to empower Micro, Small, and
Medium Enterprises (MSMEs) across Africa. Through a single integrated platform, Fanaka combines
ethical, cash-flow–based loans, financial literacy training, and microinsurance, built
specifically for the informal, women- and youth-led enterprises that traditional banks cannot
serve.
1,600+
MSMEs served
3,030+
loans disbursed
95%+
repayment rate
Challenge
Replacing manual, spreadsheet-based credit scoring at scale
Before implementing GiniMachine, Fanaka's lending operations faced two key challenges.
- Borrower data was underutilized. Rich customer information sat unused for marketing and business development, limiting Fanaka's ability to tailor offerings to its customer base and act on the data it already held.
- Credit scoring was entirely manual. The process ran on Google Sheets, it was slow, hard to scale, and unable to flag when an otherwise good customer was likely to default. The manual approach consumed substantial staff time, created missed opportunities, and undermined the accuracy of credit portfolio reporting.
Approach
AI-powered credit scoring and data-driven decisioning
To replace its manual process, Fanaka worked with GiniMachine to build, validate, and deploy an
automated credit scoring model trained on its lending history. The rollout followed a clear
sequence of steps:
-
Consolidating and preparing borrower data
Fanaka's borrower information, previously scattered across Google Sheets and underused, was brought together into a single structured dataset, then cleaned and prepared for modeling. This turned idle data into a reliable foundation for training, capturing the cash flow and behavioral signals relevant to MSME borrowers. -
Building and validating the scoring model
Using GiniMachine's machine-learning engine, Fanaka built a custom model from its historical repayment outcomes. It learned the patterns that separate likely repayers from likely defaulters, including the early-warning signals a manual review tended to miss. The model was then validated against proven, Gini-based performance metrics to confirm it was accurate enough for live decisioning. -
Automating live decisioning
Once validated, the model was deployed to score incoming applications automatically. With a single action, Fanaka can now run a full risk assessment in place of the manual, sheet-by-sheet review, cutting time-to-decision and standardizing how every applicant is evaluated. -
Customizing and scaling
GiniMachine let Fanaka tailor its scoring criteria to the MSME segment and scale the process as application volume grew, so consistent, data-driven assessment held up across a rising number of borrowers rather than breaking down under manual effort.
Result
Automated scoring and sharper portfolio insight
Implementing GiniMachine produced measurable operational gains across Fanaka's lending workflow:
- Credit scoring was automated end-to-end, removing the manual Google Sheets process and reducing time-to-decision.
- Portfolio reporting accuracy improved, enabling better-informed lending decisions and risk management.
- AI-powered analytics surfaced insights into customer behavior, supporting predictions of customer lifetime value and churn.
- The risk of missed opportunities dropped as scoring became consistent and repeatable.
- Strengthened scoring capacity helped Fanaka extend financial inclusion to more MSMEs with a focus on women- and youth-led businesses in Zambia, reaching company-wide figures of a 95%+ repayment rate across 3,030+ loans to 1,600+ MSMEs.
Download PDF version
