Raiffeisenbank implemented speech analytics of group of CST

2020: Transition to speech analytics of CST

Raiffeisenbank reported on December 2, 2020. that together with group of CST completed process of transition to upgraded version of speech analytics for quality control of service in contact center. Use of speech analytics allows to optimize processing of service addresses, to reveal painful points in dialogs and to react quickly to growth of addresses on certain subjects. Thus, the bank can increase customer service quality in remote channels and increase a share of the addresses solved at the first contact.

The speech analytics works based on own product of CST SmartLogger. In contact center calls register and are automatically transferred to the text in a format of the client operator dialog with per second timing. Also for the analysis the text addresses which arrived from the channel of chats are imported. The speech analytics of dialogs and chats allows to analyze interactions on large volumes and the periods in several hundred thousands of contacts without attraction of additional resources. Not only lexical components of dialogs, but also different attributes of records are analyzed. In addition to quantitative parameters of phonograms, such as share of the speech of the client, share of the speech of the operator, number of hecklings etc., special parameters for a chat dialogs are considered: time to the first reply of operator in a chat, the maximum time of the reply of operator and others.

Determination of subjects with the tightened answers, identification of lexical and quantitative markers of long service allows to reveal problem zones and to improve quality of client support. Also a system allows to control observance of standards — politeness, lexicon, rules of installation on Hold and other, directionally to train operators. For example, the built-in function of search of a key word and phrases in speech section after certain phrases gives the chance to analyze work with objections and failures of clients, and the lexical statistics when comparing selections reveals the best practices among unique phrases of successful dialogs.

Remote channels of service allow us to be always where it is convenient to the client, and technologies of automation – to give the answer instantly and precisely, saving comfort of live interaction. Thanks to technologies of speech analytics we can increase quality and accuracy of automatic answers and to reveal problem places in dialogs always to maintain high quality of customer service — the chief executive of department of remote work with clients of Raiffeisenbank Irma Metreveli comments.

Technologies of speech analytics helped us to be prepared for growth of addresses on a hot line at the beginning of a pandemic: in the first days of splash we analyzed the most popular subjects of addresses, parts of questions and adapted answers of the chat-bot for the most effective work with the entering addresses. Thus, in March we could close every third message with participation of the chat-bot completely automatically, without involvement of the person — the head of development of the channel of chats and the chat-bot of Raiffeisenbank Ilya Shchirov notes.

In the solution the speech sensing technology created in CST which allows to distinguish a natural spontaneous human speech with its dynamics and emotions is used, including — when interlocutors interrupt each other, speak at the same time and is not legible. So a system allows to transform to the text, and then to analyze even those dialogs in which clients spoke in noise is not quite legible. In it is the key difference of the solution of CST of the majority of the solutions presented at the market. We will continue the development for providing modern technological solutions to the financial industry — the CEO of CST group Dmitry Dyrmovsky comments.