CyVoice Analytics Admin Guide

The Voice Analytics user interface can be accessed from your web browser (MS Edge and Chrome recommended). This means that it can be accessed from anywhere as long as there is a link between the user's system and the server hosting the application.

To navigate to the portal, simply type the URL into the address bar as you would for any other webpage. The address will typically be in the following format, where HOSTNAME would be replaced by the servername, IP address or SSL Certificate of the CyTrack system. 

https://HOSTNAME//cyVoiceAnalytics/login/ 

Upon navigating to this address the main login splash screen will appear, asking for two fields of data. Your username and password. The default account is Administrator, and the secure password setup during implementation. Log in with the Same administrator credentials as used for CyReport.

The home and Status tab displays a list of recent Analysed calls and their state.

• Id - Unique ID for the job completed.

• Submitted - The date the voice analysis was submitted.

• State - Status of the job.

• Last Status - The dates the voice analysis was completed. 

• Message - Comments and summary details of the call that was analysed. 

The system configuration allows users to setup the primary language to be used for the call transcription and analysis.

• Voice Service - Third party service used for transcription.

• Speech model language - Language to be used in the call transcription.

• Profanity Filter Mode - Filter used to handle profanity.

• Azure Speech Service Endpoint - Web service that a client has access to.

• Model - The engine model used to perform the work.

• Authentication Token - The token to be used to connect the call analytics subscription. This depends on the voice provider such as Microsoft. This token will need to be organized through Cytrack.

• Language Authentication Token - Token-based authentication

• Language Service Endpoint - Supported Cognitive Services endpoints.

• Azure Storage Account Container Name - Storage blob where recordings are transcriptions are stored.

• Temporarily use Azure Storage for recordings hosted elsewhere - temporary storage blob where transcriptions are stored if needed.

The profanity filter allows users to filter inappropriate words in transcriptions. The filter allows three options:

• None - No filtering is applied and words are displayed unchanged.

• Masked - This option replaces the inappropriate word with asterisks.

• Removed - The inappropriate word is removed from the transcription altogether.

Targets

Targets allow configuration of which calls are to be analysed. This can be done against specified extension, agents, queues or campaigns.

• New Target - Allows users to configure a new target set that can be applied to keyword groups.

New Target

Clicking on the New Target button allows users to configure and select extensions, agents, queues and campaigns to be used in the call analysis. Multiple targets can be specified. For example, this would enable monitoring of calls on one group of agents for Name specific keywords, while calls through a separate queue are monitored for mentions of designated colours. Multiple keyword groups can be linked for a specified target group.

• Accept edit and Add - Saves the new configuration selection for Targets.

• Cancel - Cancels edits made in the target configuration.

• Delete - Deletes the target configuration set without confirmation.

• Name - Target name that will appear on the left-hand plane list. 

• General - Checking this option will apply the target on all calls. This means further selection of extensions and queues is not necessary.

• Speed Boost - This option decreases the time take to complete the analytics calculations. Additional cost may apply when using this option depending on the supplier.

• Linked Keyword Groups - List of available Keyword Groups that can be linked to the Targets for analysis.

• Transcribe calls on - Dropdown list of Extensions, Agents, Queue and Campaigns that can be selected for the targeted data set.

Keyword Groups

Keyword groups allow configuration of related keywords to be flagged in analysed calls. For example, a customer looking to monitor calls for specific names or colours might configure two keyword groups to classify these.

• New Keyword Group - Allows users to configure a new keyword Group that can be applied to keyword groups.

New Keyword Group

Clicking on the New Keyword Group button allows users to configure keywords to be analysed in calls. Keywords can be categorized under Keyword Groups and assigned to applicable targets.

• Accept edit and Add - Saves the new configuration selection for new keyword group.

• Cancel - Cancels edits made in the new keyword group configuration.

• Delete - Deletes the new keyword group without confirmation.

• Name - new keyword group name that will appear on the left-hand plane list. 

• Send email on keyword match - enables email alerts for keyword within the Keyword Group.

• Email Address - email address for keyword alert notification

• Report Id - Select report to be sent in email for keyword alert.

• Require all keywords to match - Ensures all keywords in the keyword group is matched for alert to be triggered.

The manual uploading tab allows uses to analyse historical calls records based on call logid.

• Upload logid - This is the ID of the historical call found in the Cytrack database.

• Id - Unique ID for the job completed.

• Submitted - The date the voice analysis was submitted.

• State - Status of the job.

• Last Status - The date the voice analysis was completed and this is the last time the system received a message on its status. 

• Message - Comments and summary details of the call that was analysed including logid and Record ID from the Cytrack database.

Click on the icon on the top right corner of the screen and click on logout to log off the Voice Analytics program.