How the largest Ukrainian bank process Big Data

PrivatBank is the largest bank in Ukraine. It serves more than 19 million customers. Here’s everything you need to know how the bank processes customer data.


Data sources

The bank analyzes a huge amount of data: from geographical coordinates and routes to customer preferences on social networks. The main sources of information collection include the following:

  • POS-terminals, ATMs and self-service terminals. They allow you to track the priority routes on which the client goes most often.
  • Accounts in social networks. If a user logs into a banking widget through social networks and agrees to give the bank access to the account (for example, to raise the credit limit), then there will be more data about him. You can find out with whom he communicates, what kind of content he likes. Based in this information, it’s possible to use predictive analytics (predict customer behavior) and identify unscrupulous customers.  
  • “PrivatMarket”. This is the e-commerce platform from Privatbank where individuals buy goods, and companies make purchases. The bank analyzes customer preferences in order to understand which products to offer.


Analysts’ work

Bank analysts decide which of the collected data will be predictors, i.e. will be important for assessing the creditworthiness of borrowers. This data is analyzed using special software based on artificial intelligence, then scoring is carried out.

For example, a credit history of people from the environment of a borrower can become a predictor. Data about this could be taken from social networks. If the client’s friends and acquaintances do not repay their debts, the bank believes that he will do the same. Moreover, even musical preferences are important. Privatbank has found that lovers of chanson worse repay debts than fans of jazz and light music.


Economic effect

It’s difficult to assess the overall economic effect. However, there are some specific cases. For example, due to accurate analysis, it was possible to additionally attract UAH 1 billion  of deposits (about $ 40 million).


Future plans

The bank is testing a new platform based on artificial intelligence together with Visa. According to Vera Platonova, Director of the Visa regional unit in the CIS and South-Eastern Europe, banks and the business they serve will use the platform.

“The platform will analyze the number of transactions, their volumes, time, place and other parameters that will allow banks and businesses to determine the segment of customers for which the provision of discounts or other offers will be most attractive. For example, a trading network, having received data on various segment groups, will know which category of customers is best to offer a discount and when, ”she explained.