Airvantage is an international micro-lender operating across 3 continents and granting over 1 million loans per day.

The Problem

At Airvantage, a credit decision must be made in real-time and at scale. Airvantage has a

highly refined rules-based system developed by its experienced credit team. Nonetheless,

Airvantage wanted to see if there was a better way to profile customers in order to reach

more customers and lower bad debts.

Our Solution

• Elucidate developed a high-performance machine learning system trained on billions of

transactions. This system uses the behavioural data of the user to inform a credit decision.

This machine learning system was used to overhaul the existing credit system in order to

inform credit decisions in real time.

• The benefits of using a machine learning solution are that machine learning algorithms are

significantly more efficient at mapping input variables to output variables than humans are.

As such, machine learning algorithms are far more capable of mapping specific user

behavioural patterns to a person’s capacity to pay back loans with significant gains in

accuracy.

The Results

• The implementation of our solution, which makes decisions in less than 0.2 seconds, has

resulted in a 27% increase in revenue at a reduced bad debt. This implies that our solution is

capable of giving loans to more people with improvements in payback capacity. This means

that our machine learning system is capable of selecting better people at a higher rate than

the previous system.

• This system therefore facilitates financial inclusion, allowing access to people who

previously had no access to micro-credit.

‍