Personalized Digital Customer Services for Consumer Banking Call Centre using Neural Networks

Abstract

Many banks and financial service companies have been transforming the way to run their businesses and serve customers. In this paper, we present a use case for digitizing customer journeys in the area of consumer banking call centre. The main objective is to provide personalized customer service experience through an integrated solution for call centre, including the Interactive Voice Response (IVR) system, SMS system, Internet Banking platform and chatbot. Topic modeling was performed on the dialogue transcript between the customers and Customer Service Officers (CSOs) to identify the customers’ reason for calling. Using the customer-level profile, transaction and servicing log data, a multi-task neural network was trained to predict if a customer is going to call the bank for any customer service request in the next 10 days. In the IVR system, a personalized voice prompt will recommend relevant digital services based on the model prediction and redirect the customer to digital services through a SMS with a URL to chatbot. Through this, the number of calls reaching the CSOs has reduced and the bank can achieve considerable operational cost savings and provide a more efficient customer service experience.