Raiffeisenbank Implements Speech Analytics from Speech Technology Center to Streamline Remote Services



2 December 2020, Moscow — Speech Technology Center and Raiffeisenbank have successfully migrated to a new version of speech analytics to monitor the quality of call centre services. The solution provides 24/7 assessment of the volume and thematic distribution of the centre's hotline calls and chat channels, enabling the transition from selective analysis of dialogues to an analysis of the entire flow of interactions for a specific period.

Speech analytics make it possible to streamline the processing of service calls, identify problem areas in dialogues and respond promptly to rising numbers of calls on specific topics. As a result, the bank can improve its quality of customer service in remote channels, increase the number of first-call resolutions and reduce the number of repeat calls. The project is unique for its integration of speech analytics with the bank's chat platform. Speech Technology Center's specialists created a unified system and a single interface for analysing not only voice messages from the phone channel but also text messages from the chat channel.

The team at Speech Technology Center implemented the speech analytics using its Smart Logger product. Calls to the centre are recorded and automatically converted into text in the format of a "client-operator" dialogue with a second-by-second breakdown. Text messages from the chat channel are also imported for analytical purposes. Applying speech analytics to dialogues and text chats makes it possible to analyse communications across large volumes and periods of hundreds of thousands of calls. Up to 100% of calls are analysed without additional resources. The system is also capable of conducting analysis using not only the lexical components of dialogues but also various other characteristics of the recordings. In addition to common quantitative parameters such as the ratio of customer speech to that of the operator, the number of interruptions, etc., the analysis also examines special parameters for chat dialogues such as the time taken until the operator's first response, their maximum response time, and more.

During the implementation stage, the Speech Technology Center and Raiffeisenbank's analysts configured the system for future activities, including the setup of analytical queries and dictionaries for call topics. The ability to measure FCR (First Call Resolution, the percentage of issues resolved during the customer's first call) for each call topic allows call centre operators to identify problem areas. Identifying topics with protracted replies and the lexical and quantitative markers of prolonged service makes it possible to reduce the average length of dialogues, making them faster and more efficient. The system can also help to improve quality of service by easily monitoring compliance with standards such as politeness, vocabulary, the rules for putting callers on hold, etc., and to provide targeted training for operators. For example, the in-built feature for searching keywords and phrases in speech following specific phrases enables operators to analyse how customer objections and refusals are handled, while the generation of lexical statistics when comparing samples helps to develop best practices for using unique phrases in successful dialogues.

'Remote service channels allow us to always be where the customer needs us to be, and automation technologies let us respond instantly and accurately without losing the human touch of live interaction. With speech analytics technologies, we can improve the quality and accuracy of automatic responses and identify problem areas in dialogues to ensure a consistently high quality of customer interactions', said Irma Metreveli, Executive Director of the Remote Customers Interaction Department at Raiffeisenbank.

'Speech analytics technologies helped us to prepare for the growth in hotline calls at the start of the pandemic. In those early days following the sudden increase in calls, we analysed the most popular topics and the details of the questions asked and adapted the chatbot's responses to handle incoming calls as effectively as possible. As a result, in March we were able to process one in three messages automatically using the chatbot, without any human involvement', added Ilya Shchirov, Head of Chat Channel and Chatbot Development at Raiffeisenbank.

'This speech recognition technology was named best in the world at the international CHiME 2020 competition. The Speech Technology Center's speech recognition technology can process natural spontaneous human speech with all its dynamics and emotions, including situations when participants interrupt each other, talk at the same time or speak unintelligibly. The system can convert speech to text, and even analyse dialogues in which the customer is speaking in a noisy environment or in a difficult-to-understand way. This is the key difference between the Speech Technology Center's solution and most others on the market. We will continue to develop in order to provide the financial sector with advanced technology solutions', said Dmitriy Dyrmovskiy, CEO of Speech Technology Center.

Speech Technology Center (part of the Sberbank ecosystem) is a global developer of intelligent speech and face recognition technologies and an expert resource in AI and machine learning with 30 years of experience in the field. 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 occupy leading positions in global rankings by NIST, ASVspoof Challenge, and VOiCES. CHiME 2020 named Speech Technology Center’s speech recognition technology the best in the world. Speech Technology Center has used artificial intelligence to implement more than 5,000 projects in 70 countries.
Media contact: pr@speechpro.com

Raiffeisenbank is a subsidiary of Raiffeisen Bank International AG. Raiffeisenbank is one of the most reliable Russian banks. It creates financial solutions for retail and corporate customers, including both residents and non-residents of the Russian Federation. According to Interfax-CEA, in the first 9 months of 2020, Raiffeisenbank ranked 10th in terms of assets, 7th in terms of funds from retail customers and 12th in terms of retail loans. In 2020, Forbes magazine named Raiffeisenbank the most reliable and best bank in Russia.