The problem: Micro-credit institutions, that work among others in developing countries, are facing many fraud attempts. This impedes both their ability to quickly deliver credit to those in need, but also how much they can lend.

Our Impact

Bayes Impact worked with micro-finance organization Zidisha, which directly connects lenders and borrowers, without going through an intermediary structure. Zidisha works mainly in Kenya and Ghana.

Bayes Impact has developed algorithms for Zidisha to quickly detect possible fraud attempts.

Following our intervention, the fraud has dropped by 35% on the platform.

Better fraud detection has led to a 10% increase in new loan applications accepted.