Sabina, a bilingual robot created on the Speech Technology Center's platform, has helped to reduce the workload of a call centre operated by Sberbank in Kazakhstan by up to 26%



14 April 2021 — Moscow — The Speech Technology Center (part of the Sberbank ecosystem) provided its Customer Engagement Platform to Sberbank in Kazakhstan to create smart robots powered by artificial intelligence.

Sabina, a bilingual robotSberbank Kazakhstan created Sabina (SABI — System Artificial Bank Intelligence), a robot designed to speed up the operations of the call centre and improve the quality of service. This voice assistant was launched in the summer of 2020 and, over this period, reduced the workload of the call centre by up to 26%. The robot advises customers across telephone and text channels on lending, money transfers, payment cards and deposits. A special feature of Sabina is its bilingualism, as the robot can communicate in both Kazakh and Russian. The robot was launched relatively recently, and its skills are currently being improved. This project is unique as it involves the use of speech analytics, voice and text robots, and biometrics, three products developed by the Speech Technology Center and powered by artificial intelligence, neural networks, and machine learning.

The key idea behind implementing the robot was to establish a smart call centre managed using artificial intelligence, and transform traditional customer service mechanisms into a multifunctional system that enables an individual approach to each customer by "hearing his or her voice".

The development of Sabina includes three evolutionary stages: the Advisor-Navigator provides the customers with general information on banking products and services and, if necessary, redirects the customers to the required group of operators. This stage has already been implemented. The Service Manager performs the functions of an advisor and provides services involving the systems of the bank, such as voice biometrics, card blocking, loan application processing, payment management, transfers, etc. This feature is currently in the active stage of implementation. The third and final stage of Sabina's evolution will be implemented in the future. It involves the Personal Manager, when the robot begins to treat the customer experience, behaviour model and the preferences of the customer as its main paradigm and motivating factors for interactions, bringing the bank's service and communications with its customers to a whole new level.

Yernar Nakisbekov, Deputy Chairman of the Board, Sberbank Kazakhstan: "The voice robot has allowed the waiting time at the bank's call centre to be reduced and the channels for simultaneous customer service to be expanded. Currently, Sabina assists retail and corporate customers in voice and non-voice service channels, including our mobile apps, such as Sberbank online and Sberbusiness. It not only advises customers on banking products but also provides other services, including card blocking and providing information on currency exchange rates. There has been a significant increase in the share of customers who used the voice robot's services. This was enabled, in particular, by the quality of synthesised speech powered by the solutions from the Speech Technology Centre, as the robot's voice is hardly distinguishable from the voice of a human operator. These solutions allow the robot to quickly explore the digital landscape and instantly provide users with relevant responses. The key benefit of collaboration with the Speech Technology Center is the ability to integrate a whole stack of AI solutions from a single vendor with a wealth of experience in enterprise-level projects: this ensures the quality of the products and their ability to smoothly integrate with one another".

Dmitriy Dyrmovskiy, CEO of the Speech Technology Сenter: "The omnichannel platform we created enables three-way interactions between the customer, the robot and the operator. Sabina's advantage lies in the prospects for integration with customer speech analytics, lead generation and auto-call systems, which will allow the robot to generate a customer profile based on prior communication experience and will enable personalised communication based on a knowledge of customer preferences. The robot can use the answers collected for self-training to improve each subsequent dialogue. With the Customer Engagement Platform, we help the managers of large call centres and IT directors cut costs in their operations, improve the quality of customer service and increase sales.

Our algorithms allow speech to be recognised and synthesised, 100% of customer-operator dialogues to be analysed, the key reasons for calls to be identified and 90% of standard questions to be answered automatically, without the need to involve any human resources. This is demonstrated by the project implemented in partnership with our colleagues from Sberbank Kazakhstan".

About Speech Technology Center: Speech Technology Center (part of the Sber ecosystem) is a global developer of intelligent speech and face recognition technologies, and an expert resource in AI and machine learning, a member of ESSMA. Speech Technology Center is one of the few companies in the world that creates and develops both biometric modalities: face and voice. The voice falsification detection and speech recognition solutions developed by Speech Technology Center score leading positions in global rankings by NIST, ASVspoof Challenge, VOiCES. At CHiME Сhallenge 2020, its speech recognition technology was recognized as the best. Speech Technology Center’s solutions are in demand in 70 countries around the world.

Media contact: pr@speechpro.com