Neurodata Lab

Neurodata Lab has developed emotion recognition technology for the bank

ROSBANK Public Joint Stock Company has launched a pilot project based on neural networks technology. The task of the new product is to recognize the emotions of users when contacting the call-center. The advantage of the platform based on the work of artificial intelligence with machine learning elements is that speech features are recognized at semantic and voice levels. To determine the vocabulary of the person who has applied to the bank, the speech is transformed into text.

The program also analyzes the behavior of the operator: pauses in his speech, raising and lowering his voice and dialogue time. As a result, managers responsible for the operation of the contact center and the general user experience can compile summary statistics on the work of the support team. The task of the analyst is to compare the received information with a growing or falling service satisfaction index (as well as a banking product). Currently, the role of services based on biometric data, big data and artificial intelligence is growing. Therefore, the introduction of this technology can take service to a new level. In the future, the solution for determining the emotions of customers of the bank can be scaled to branches — to obtain statistics of communication of employees with users.

Neurodata Lab platform and its integration into the CRM-system of a financial institution

Thanks to the integration of the platform developed by Neurodata Lab into the CRM system, responsible managers have a complete picture associated with each client, operator, bank employee and office. The second stage of development of this project is analysis of streaming video from video surveillance cameras directly in the department. This will made real-time analytics about the factors that form customer experience available. Also, marketers will be able to track the reaction of users to new consumer services. Not only the above-mentioned financial institution, but also Alfa-Bank interested in studying clients’ emotions. In one of the offices, the expression of the visitor’s face is currently being analyzed. From the received data the conclusion about the general satisfaction of the user with the service is made. This technology is designed to receive feedback without callbacks and polls.

In order to interpret the emotions received from the video stream, our own developments are used by the technical specialists of Alfa-Bank. Employees consider the found dependence of customer emotions on the quantity and quality of products sold as the main success of innovation.

The system, which unites 160,000 video cameras and integrated with the neural network, is successfully operating in Moscow. The main purpose of the platform is to recognize persons and find offenders. Also in Europe the system iBorderCtrl is popular on border control. Analyzing the mimics of the tourist, the system determines whether a person who crosses the border tell truth or lie. This platform is combined with a complete biometric base of people who have ever entered the EU.