Speech analytics service with YandexGPT is now available in Kazakhstan

Yandex Cloud has introduced the Yandex SpeechSense speech analytics service. This technology will enable companies in Kazakhstan to conduct detailed analysis of operator calls and chats in contact centers, including evaluating their tone and engagement in conversations. With the new system, businesses can improve service quality in call centers and reduce costs associated with analyzing employee-customer communications. The service is available to platform users in Private Preview mode and is provided upon request.

SpeechSense speech analytics involves several stages:
— first, call recordings, chats, and data from CRM systems are uploaded to the cloud

— in the cloud, speech is converted into text using Yandex SpeechKit's speech recognition technology

— the resulting textual content undergoes automatic verification for compliance with scripts. For example, it is marked whether the operator greeted the customer or if there were any negative statements in their speech

— after that, the transcript is sent for more comprehensive analysis in Yandex GPT, which is integrated with the service.

Integration with Yandex GPT provides greater capabilities for speech analytics in the service. Currently, SpeechSense uses a general neural network model. Developers of the cloud platform plan to continue training and improving it so that it can, for example, check operator responses for compliance with the service's editorial policy and business processes.