Sber has reduced the routing time of business clients to the contact center operator by 3.5 times

2023

Savings from the introduction of AI for the year amounted to 300 million rubles

In 2023, savings from implementation artificial intelligence contact center Sberbank for legal entities amounted to 300 million. rubles Sberbank announced this on February 29, 2024.

The average time for routing calls from Sberbank corporate clients to the support line in two years has decreased by 3.5 times and as of February 2024 is only 18 seconds. The virtual AI assistant predictively routes the requests of 87% of clients. Thanks to the use of AI technologies, routing accuracy increased to 77% over the year, which is 14 percentage points. more than the indicator without the use of AI technologies.

In the text channel, the AI assistant independently distributes requests between operators, analyzes the quality of consultations and successfully solves 23% of requests. In the voice channel, voice analytics already covers 99% of consultations. Artificial intelligence analyzes the course of the dialogue, assesses the quality of the consultation and the operator's compliance with customer service standards.

Artificial intelligence is also used to predict CSI - an ensemble of machine learning models predicts this figure for 100% of referrals. And the operators of the contact center now have a personal AI mentor - co-pilot, which works in the background and is always ready to answer the most difficult question.

Шаблон:Quote 'author = said Sergey Lekhanov, Director of the Corporate Solutions Center Division of Sberbank.

Increase by 20% in the resolution of issues of corporate clients of Sberbank when first contacting the contact center

Sberbank on February 1, 2023 shared the results of the implementation of projects using artificial intelligence (AI) technologies in a contact center for legal entities. At the end of 2022, the introduction of AI into the work of the chatbot increased the number of customer questions resolved during the first appeal by 20% year-on-year.

AI technologies are used in the work of the contact center for speech recognition and analysis, classification of calls by a virtual assistant, as well as to automate the process of verifying document scans. As a result, the quality of service for customers is improved and the volume of routine processes for support line operators is reduced.

Шаблон:Quote 'author = said Sergey Lekhanov, Director of the Corporate Solutions Center Division of Sberbank.

2018: Robotization of the contact center for corporate clients

In Sberbank February 2018, he began robotizing contact center for corporate clients. ATMs bank robot Anna will answer questions about the location with the function of self-collection and branches to clients. It is planned that in the future Anna will provide other reference information: for example, on the status of payment, account balance, on executive lists.

According to the bank, reference requests account for approximately 40% of calls to the Sberbank contact center for corporate clients. Its operators process about 20 thousand calls every day. Requests that require the participation of an employee are also processed using robotic technology: using a cognitive virtual assistant. The robot analyzes the conversation between the operator and the client and independently "from the voice" provides the operator with additional information necessary to provide advice.

According to Sberbank estimates, according to the results of the pilot project, the speed of customer service on certain topics has increased by 50%: now the average length of a client's call to a contact center for corporate clients is 3.5 minutes.

We use artificial intelligence to solve a wide variety of problems, and one of them is to reduce the service time for our clients due to high-quality and convenient service achieved by automating the work of the contact center, "said Anatoly Popov, Senior Vice President, Head of Corporate Business at Sberbank. - Sberbank's goals are to significantly reduce the average customer service time, increase the share of solving customer requests at the FCR (First Call Resolution) stage, the value in the context - at the first customer contact, note TAdviser) and transfer all simple customer requests to the robotic processing system within two years.