Sberbank of Russia : Chat bot Sber designs for Moscow Metro handles 88% of queries without human operator

December 3, 2020, Moscow - Speech Technology Center Group, a Sber ecosystem resident, Sberbank's Moskovsky Bank, and Moscow Metro have introduced a chatbot for passengers. During a focus group testing, the virtual text assistant automated 75% of requests without putting calls through to a human operator and was put into commercial use on November 17. Being used commercially, the virtual assistant delivers an automation rate of 88%. The bot is available in the Moscow Metro application.

Roman Latypov, first deputy metro head for strategic development and customer relations: 'Our metrobot is becoming one of the most technologically-advanced and modern information services for customers. It performed great during its very first days on the job: the help of a human operator was required in only 12% of cases as these requests were too long or didn't have clear wording. The chatbot features dialogue scripts for more than 17,000 questions. Alexandra - this is the name our passengers have chosen - can talk about fares, station and transfer timetables, and not only for the metro but also for MCC and MCD. And this is only the startup, we will expand this list down the road.'

Gartner predicts that by 2024 nearly 70% of routine work will be automated with AI, including chatbots. The virtual assistant STC designed for Moscow Metro has an 88% automation rate already today.

The STC team has created the chatbot based on its proprietary ChatNavigator, an omnichannel dialogue platform for creating intelligent virtual assistants. A universal connector (API) has been embedded in the chatbot, which can communicate with passengers across channels without the help of an operator. The basic communication scenario remains the same for all channels, but each considers its communication peculiarities.

The most popular topics of inquiries the bot received during the focus group testing were fare policy and fare payment methods, travel concessions, metro services, and general questions about the quality of services provided.

Dmitry Dyrmovsky, CEO, STC Group:'The project is unique as all components of the chatbot and platform are in the cloud. This is one of the first 100% cloud-based solutions marketed. To develop the bot and understand the meaning of queries more accurately, hybrid classification technology is used, combining rules and machine learning. This will allow creating new topics and developing current ones, improving the virtual assistant continuously.'

Vyacheslav Tsybulnikov, Vice President of Sberbank, Chairman of Moskovsky Bank: 'Sberbank offers Muscovites and guests of the capital cutting-edge digital services in finance and many other areas of everyday life. We are confident that the chatbot, which was developed with the participation of our subsidiary STC, will make communication with passengers more efficient and comfortable not only in the metro but, moving forward, in other types of public transport too.'

STC Group, a Sber ecosystem resident, is a global developer of products and solutions based on intelligent speech technologies and face recognition. A technology expert in artificial intelligence and machine learning with 30 years of experience, STC is one of the few companies in the world that creates and develops both biometric modalities, face and voice. STC Group's detection of voice forgeries and speech recognition hold top spots in the world rankings by NIST, ASVspoof Challenge, and VOiCES. At CHiME 2020, STC's speech recognition technology was recognized as the best in the world. STC Group has implemented more than 5,000 projects using artificial intelligence in 70 countries.