Deploying and configuring Verint transcription

Verint transcription is using the Verint engine to transcribe voice calls. The engine is built into the Verba Speech Analytics Service which can be enabled on the Verba Speech Analytics Server role. The Speech Analytics Server requires special sizing when the Verint transcription engine is used, for more information see Deploying transcription.

Once, the server(s) are deployed, the required Data Processor has to be created to enable the integration with the Verint transcription engine. Follow the steps described in Configuring and running transcription to create the processor and select the Verint engine. No additional settings are required for the data processor configuration.

After creating the data processor, you can follow the guidelines at Configuring and running transcription to configure one or more data management policies.

License requirements

The Verint transcription engine requires specific licenses to be purchased and uploaded to the system. Please contact your Verint sales representative for more information.

After uploading the necessary licenses, the licenses have to be assigned to users through the role configuration. The following permission(s) are required for the Verint speech transcription:

• Transcribe Conversations (Built-in)

Uploading new ASR models

Follow the instructions below to upload an ASR model (often referred to as language models) to the system.

Copy the language model file, which is a .zip file, to C:\Program Files\Verba\resources\transcription\eliza to every server where the Verba Speech Analytics Service is enabled. For additional languages, only this step needs to be repeated. The service will detect the new language model within 30 minutes, or you can force the detection by restarting the service. After the service detects the new model, it will be listed on the speech transcription data management policy configuration page.

Customizing the Verint ASR models

Please read the Improving transcription accuracy for Verint Speech Transcription for more information.