Sber has implemented more than 200 artificial intelligence models that help the company create customer profiles, plan, control and account, and make decisions on debt repayment. Kirill Demin, Deputy Head of Sberbank’s Department for Working with Troubled Assets, talked about this issue at the conference “Collection Technologies in the Age of Digital Transformation” held in Samara.
According to him, digital technologies in typical clearing and collection processes allow them to be carried out faster and at lower costs. The sooner a debt can be paid off, the greater the variability of payment and the lower the debt amount. It also increases the likelihood that the property will be sold for a higher price if needed.
Kirill Demin stated that the digitalization of layouts reduces the likelihood and importance of errors for employees.
“In retail reconciliation, our models allow us to choose the right tactics to engage the customer; whether he needs to make a call, send an SMS, arrange a personal meeting or go to court. We have long been using an artificial intelligence-based robot operator to communicate with customers – this allows us to make 84% of calls without human intervention. The economic impact of using robots is estimated at 2.4 billion rubles per year,” said the director of the Sberbank department.
Digital transformation in enforcement proceedings means the introduction of electronic enforcement documents, their routing through digital channels and the digitalization of the registry enforcement system.
Demin added that Sber has almost 400 thousand electronic enforcement letters worth 78 billion rubles, of which 320 thousand are notarized enforcement letters (about 50 thousand enforcement letters).
According to the top manager, thanks to the transition to electronic enforcement documents, the time to initiate enforcement proceedings has been shortened by 10 times and today it does not exceed 6 days.