The challenge
In Mongolia, traditional credit scoring kept large parts of the population out of the financial system. Two problems stood in the way.
Manual processing. Every application was scored by hand. That made lending slow and expensive, and it made small loans impossible to offer profitably, so lenders could not scale.
Narrow data. Credit decisions relied on credit history, proof of steady income and collateral. People without a credit history, those working in the informal economy and anyone without documented income or assets were effectively shut out.
The solution
LendMN launched with Mongolia's first AI/ML-powered credit scoring system. Multiple machine learning models, including random forests and artificial neural networks, score each applicant automatically and keep learning as more loans are repaid. The whole journey, from application to disbursement, runs on a smartphone.
As the customer base grew, LendMN added behavioural scoring. It uses repayment patterns and in-app activity to spot high-value customers, offer product upgrades and set personalised credit limits.
The results
- 96% repayment rate, well above the roughly 80% average at traditional financial institutions.
- 5-minute approvals, down from days or weeks under the manual process.
- 180,000 loans a month, making LendMN the leading micro-lending provider in the market.
Just as important, the model opened access to credit for micro-entrepreneurs, informal workers and people in remote areas, showing how AI-driven lending can grow a portfolio and financial inclusion at the same time.
.avif)


