Russia's VTB is anticipating a 30% uplift in call centre automation following the introduction of a voice assistant that can anticipate client demands.
Initially operating in pilot mode, the VTB voice assistant will not only identify the client at the time of the call, but also analyse their history of interaction with the bank, including previous requests. The system will form a suggestion of what the current query is related to, and immediately after the greeting, it will clarify the request with the client, for example, the amount owed on a loan.
If the client confirms that the query is related to the predicted topic, the voice assistant will immediately provide an answer, or direct it to an operator. Alternatively, the client is asked to share their query with an operator, and the algorithms of the predictive model are adjusted.
Maxim Konovalikhin, SVP and head of VTB data analysis and modeling department, comments: "Thanks to the introduction of predictive models for the voice assistant, we will be able to more accurately predict the wishes of our customers, and promptly answer their questions. In the future, the specialised infrastructure built by us for this solution will allow us to solve a whole range of new queries instantly providing a response to our customers’ needs through the use of AI models."
In 2021, VTB increased the number of implementations of voice assistant services in the contact center on a variety of queries by 7.5 times. Advanced intelligent routing capabilities have also reduced the number of switches when dealing with customer issues.