Real-Time Agent Assist deployment workflow

Follow the steps in the workflow to deploy Real-Time Agent Assist in your environment.

Before you begin 

  • Verify that your organization is registered for Common Cloud Services. Registration is required for accessing RTAA cloud-based services. If your organization is not registered, contact Cloud Services.

  • Verify that WFO V15.2 2021R1 or higher is installed.

  • If your recorders are on premises, verify that there is sufficient available bandwidth for each recorder, according to the expected load of streamed audio to RTAA. For details, see Recorder Server Sizing (Performance and Sizing Guide).

Workflow 

Task

Details

Reference

Enable access to RTAA cloud services

Open a ticket to Cloud Services to enable access and assign API scope to the following services:

  • Work Assist

  • Real-Time Linguistic (remote) engine

  • Interaction Wrap-Up Bot

  • Summary and Transcription Fetch

  • Agent Transfer Bot

 

Install hot fixes

Install the required hot fixes for your RTAA solution.

Required hot fixes

Enable licenses and feature flags

For each RTAA module you choose to deploy:

  • Upload and activate the relevant license.

  • On the License Edit page (http://<servername>/wfo/
    control/license_edit    ), enable the relevant feature flags.

Real-Time Agent Assist licenses, feature flags, and server roles

Configure server roles

For each RTAA module you choose to deploy:

  1. In Enterprise Manager, activate the relevant server roles.

  2. Configure settings for the activated server roles.

Configure external server roles

For the Work Assist and Real-Time Linguistic (remote) cloud services, configure external server roles:

  1. In Enterprise Manager, add an external server for each to the installations tree.

    In a multi-tenant environment, the external server must be configured under the host organization.

  2. On the external server, activate the relevant external server role.
  3. Configure settings for the activated external server role.

  • Enterprise Manager Configuration and Administration Guide:

    • "Add an external server to the installations tree"

    • "Activate server roles on an external server"

    • "External Real-Time Agent Assist Cloud Configuration Server Role and Associations"

    • "External Work Assist Cloud Configuration Server Role and Associations"

Deploy language model

(only required with Real-Time Linguistic (remote) engine)

To deploy a boosted language model, or to transcribe interactions in a language that is not yet available in the Verint cloud, contact Business Enablement to deploy the relevant language model.

 

Allowlist URLs

Allowlist URLs for the relevant server roles.

Allowlist reference

Install Work Assist desktop client and set up users

From the Software Tools page on Verint Connect, download the Work Assist desktop client installation package and install it on user desktops.

"Work Assist client" (Desktop Applications Deployment Reference and Installation Guide)

In User Management >Employees > Profiles, configure the following settings for each user profile:

  • Desktop Messaging Username. Provide the username (email) to sign in to Work Assist in the format user@domain.com.
  • Organization. Select an organization that will be used to apply Work Assist Rules (in Tracking, under Notifications, Organization Rules).

"Employee information and privileges" User Management Guide

In User Management >Security > Usernames, for each user profile provide a valid Username and Password and set the Status to Active. The username must be in the format: user@domain.com

"Set up user accounts" User Management Guide

Enable real-time categories in RTAA rules

Real-time categories cannot be used in an RTAA rule if the language model associated to these categories is different than the language model associated to the RTAA rule.

Install the relevant hot fix, and then perform the following configuration on the CentralApp database:

USE [CentralApp];

GO

UPDATE [dbo].[Params]

SET [value] = 1

WHERE [name] = 'PRM_RTAA_IGNORE_CATEGORIES_LM';

GO

After performing this configuration, when configuring an RTAA rule, real-time categories are displayed in the category filter.

Required hot fixes

Open a ticket to Services to complete the RTAA deployment

After you complete the steps in the workflow, open a ticket to Cloud Services to complete the deployment of the language model.

 

(Optional) Contact your account representative to integrate Bring Your Own Model (BYOM) with Interaction Wrap-Up Bot

To integrate your own GenAI services, the following adapters are available:

  • API Service Adapter: Allows you to use any LLM and your own prompts.

  • Customer LLM Adapter: Allows you to only use supported LLMs with default prompts.

 

Configure a dedicated Recorder Analytics rule for Interaction Wrap-Up Bot and Agent Transfer Bot

Configure a dedicated Recorder Analytics rule for:

  • Interaction Wrap-Up Bot to generate summaries for a subset of interactions only.

  • Agent Transfer Bot to transfer summaries of a subset of interactions only.

 

(Optional) Open a ticket to Services to customize your Interaction Wrap-Up Bot

The Wrap-Up Bot can be configured to generate summaries using a custom summary format you provide. Based on the Recorder Analytics rule that is triggered, the Wrap-Up Bot selects the summary format to use.

 

(Optional) Open a ticket to Services to enable using custom attributes to fetch real-time transcriptions and summaries via an API call

You can select up to three (3) parameters from the call manifest, such as agent ID, ANI, or holds to be configured as custom attributes for fetching transcriptions and summaries via an API call. Only one custom attribute can be used per API call.

 

Related topics 

Real-Time Agent Assist licenses, feature flags, and server roles