Optional: Distribute RTSA language model

If required, distribute a language model for Real-Time Speech Analytics (RTSA) from the Recording Management application on the WFO Portal.

If the language model for regular Speech Analytics is a new language model, it is recommended to review the results before distributing the model to Real-Time Speech Analytics.

Before you begin

Verify that the language model for distribution is RTSA compatible. See the Real-Time Speech Analytics Set Up Guide, section “Obtain the Real-Time Speech Analytics language model”.
Copy the language model to distribute from %impact360datadir%\ SharedFolders\PhoneticBoosting\PBOnsiteModels locally or into a network location.

Procedure

Sign in to WFO.
Go to Recording Management. Under Recorder Analytics Rules, select Real Time Speech Analytics Models.
Select Create and do the following:
1. In the Description field, add a description that uniquely identifies this language model.
2. To make this language model the active version, select Active.
3. To upload the language model you copied, select Browse.
4. In a multi-tenant environment, from the Tenant list, select one of the following:
  - Tenant: Select the tenant to which you want to associate the Real time Speech Analytics model. When creating a Recorder Analytics Rule, if the rule is associated with the same tenant, the RTSA language model is available in the list of language models.
  - Shared: When creating a Recorder Analytics Rule, the language model is available for selection, irrespective of the tenant associated with the rule.
To distribute the selected language model, select Save. This distributes the new boosted language model from Speech Analytics to Real-Time Speech Analytics.