Rosbank began testing the neural network emotion recognition technology in the voice of the client

    14 February

Rosbank launched a pilot to analyze customer service satisfaction in real time. The system developed by Neurodata Lab will be implemented in a test mode in the contact center.

The solution based on neural network technology recognizes the emotions in the voice of customers who applied to the contact center, and calculates the satisfaction index with service directly during a conversation with the operator. At the same time, the analysis is implemented both at the voice and at semantic level, when translating speech into text. The system also takes into account the additional parameters: the number of pauses in the operator’s speech, the change in the voice volume and the total talk time. This kind of analytics allows recognizing customer emotions and highlighting the most critical classes of calls. Thus, the contact center manager will receive summary statistics on the work of operators with each call: recognized emotions, dynamics of changes in the satisfaction index with service over time, and comparative indicators of the efficiency of the provided service.

Vasily Voronov

Acting Innovation Director

Today different technologies based on biometrics are becoming more and more popular in various fields, including banking. Rosbank is working constantly to improve the customer service quality and we hope that the introduction of technology to recognize emotions in the voice will help us bring the service to a new level. In case of successful pilot project implementation the system can be used in the network of bank branches for analyzing employee communication with customers.

Georgy Pliev

Founder and CEO of the Neurodata Lab company

The Neurodata Lab solution connects the new emotion recognition technology to the market and the real demand of the banking industry. Due to the possibility of integration into the CRM system, the bank receives more detailed and deep analytics for each client, operator and the whole branch. At the same time, we can work not only with voice data, but also analyze streaming video from cameras in sales offices in real time. Such analytics will allow the bank to track factors affecting the quality of customer service accurately, including assessing how customers perceive innovations in branches. This could be the second step of the pilot project.

About Neurodata Lab
Neurodata Lab conducts research and development of multimodal systems for recognition and synthesis of emotions, analysis of social behavior, speech and movements, as well as non-invasive methods for measuring physiological signals. The company specializes in solutions for the retail, banking and insurance industry, HoReCa and service robotics.