SberDevices voice technologies have improved MegaFon's voice platform

2024: SaluteSpeech Implementation

MegaFon introduced synthesis technologies speech recognition SaluteSpeech and from SberDevices to improve her virtual assistant Elena, who is entrusted with communicating with the company's clients. This was Sber announced on April 17, 2024. As a result, in terms of cognitive parameters, the assistant came close to the operators. contact center Elena's speech recognition speed increased by 20%, and recognition accuracy increased to 95%. The tests carried out during the implementation show that with the new functionality, the level of customer loyalty increased by 3.5 percentage points.

Шаблон:Quote 'author = said Denis Filippov, CEO of SberDevices.

{{quote 'author = noted Dmitry Rudskikh, Commercial Director of MegaFon. | Elena was always empathetic and could find the necessary words for each situation in the dialogue with clients. We consistently developed and supplemented her abilities so that she sounded more human and spoke in her own voice, and at the same time resolved issues quickly and simply - this is what our client wants when communicating with an assistant. It's great that the introduction of SaluteSpeech speech technologies from SberDevices opens up new opportunities for us to develop the product,}}

The SaluteSpeech model was specially trained for the vocabulary of the mobile operator so that the assistant could better recognize client requests. In addition, Elena received a unique voice from the SaluteSpeech YourVoice catalog, which is now exclusive to her and associated with the MegaFon brand.

The implementation of the project - from signing an agreement to industrial implementation in the perimeter of the company for 100% of flows for all regions of Russia - took only three months. All work was carried out in the circuit of MegaFon and on its equipment with optimized resources for the volume that the customer initially laid down.