Case StudyScotiabank

Scotiabank improves payment collections of credit card customers using deep learning

Scotiabank is using deep learning to better manage credit card collections. The platform developed by Dessa identifies risky customers and classifies them according to risk which is calculated by analysing historical data from the bank. It then predicts whether the customer needs a reminder or not, whether there is a chance of delinquency and so on.

Context

Bank of Nova Scotia has 5 million Canadian credit card customers and wants to integrate AI into its operations

The Project

"Scotia Bank uses AI to analyze customer behavior for eg: figure out the best way to interact with customers who may have simply forgotten to pay versus those less likely to pay at all. A simple nudge via text message might be more effective than call. The DeepLearni.ng tool predicts possible delinquencies faster and suggests collection options based on the type of customer such as a forgetful payer. It analyzes individual customer behavior, looks for patterns and correlates it with other data pulled from the bank’s enterprise data lake. It then suggests the best time and method, such as email or text message, to reach credit card customers about payments."

Data

Structured transaction data

Results

Results undisclosed

Related Use Cases

Use Case

Analyse credit risk of individual customers

→

←Back to Case Studies