V3 Voice Recordings

Owned by XCALLY Knowledge (Unlicensed)
Last updated: Sept 11, 2024 by Product Knowledge
7 min read

What’s about

XCALLY allows you to record voice calls.

  • You can enable call recording for Voice Queues: calls going through the enabled queue will be automatically recorded:

    • Voice Menu → Voice Queues → Edit Voice Queue → Advanced Settings → Recording Format (to enable, choose a recording format)

  • You can enable call recording for Queue Campaign: calls going through the enabled queue campaign will be automatically recorded

    • Motion Bull → Queue Campaigns → Edit Queue Campaign → Advanced Settings → Recording Format (to enable, choose a recording format)

  • You can enable call recording for Agents: Agents can manually record calls

    • Staff → Agents → Edit Agent → Phonebar Settings/ WebRTC Settings → Enable Recording

Split Voice Recordings

XCALLY allows you to record voice calls with a single audio per speaker using the Split Voice Recording feature.
The system will record in stereo and not in mono

This is the first version of the Split Voice Recordings feature. We are still working to better your experience using this new product improvement

Requirements

  • You need to install the 'ffmpeg' component Command for Debian: apt-get install ffmpeg

  • And restart the motion service to apply the change

The Split Voice Recordings feature is available for:

  • Voice Queue Calls automatic recordings (inbound, outbound, internal routes)

  • Motion Bull Queue Campaign automatic recordings, for all dialer methods

  • Manual recordings by Agents via Phonebar or WebRTC

Please, be aware that enabling the Split Voice Recordings feature may increase CPU usage and potentially affect performance. We recommend you to monitor the system performance while using this feature.

From the Settings Menu → go to General → go to Global section and Enable audio split for voice recordings

Stop Recording on Voice Call Transfer

From General settings, you can select this option if you want the recording to stop when a call is transferred (discover how transfer works on this page).

  • For outbound calls, the recording will be stopped if you switch on the option

  • For inbound calls, stop recording on voice call transfer works in different ways, depending on whether the transfer is blind (when agent1 transfers the call to agent2 immediately) or attended (when agent1 calls before agent2 and then he transfers to them the call). In case of blind transfer the recording will always be stopped, in case of attended transfer depends on whether the transfer is internal or external.

SCENARIO

Stop Recording on Voice Call Transfer
Transfer Type
Recording
Stop Recording on Voice Call Transfer
Transfer Type
Recording
ON
BLIND to Internal
1 REC (Agent1&Client1)
ON
BLIND to External
1 REC (Agent1&Client1)
ON
ATTENDED to Internal
1 REC [(Agent1&Client1) + (Agent2&Client1)]
ON
ATTENDED to External
1 REC (Agent1&Client1)
OFF
BLIND to Internal
1 REC [(Agent1&Client1) + (Agent2&Client1)]
OFF
BLIND to External
1 REC [(Agent1&Client1) + (Agent2&Client1)]
OFF
ATTENDED to Internal
1 REC [(Agent1&Client1) + (Agent2&Client1)]
OFF
ATTENDED to External
1 REC [(Agent1&Client1) + (Agent2&Client1)]

 

Inbound Calls
Agents → Edit → Record calls to user extension →Inactive (on both agents involved), so recording is not automatic on the agent channel
Client1 calls on a internal route Agent1 (not a Queue which has own settings)
Agent1 starts the registration manually and then transfers the call to Agent2

If you switch ON Stop Recording on voice call transfer:

  • With Transfer Type BLIND to Internal/External, output will be with a call recording only between Agent1&Client1

  • With Transfer Type ATTENDED to External, output will be with a call recording only between Agent1&Client1

  • With Transfer Type ATTENDED to Internal, output will be with a call recording between Agent1&Client1 + recording between Agent2&Client1

If you switch OFF Stop Recording on voice call transfer there will always be an output with a call recording between Agent1&Client1 + recording between Agent2&Client1

Voice Recording Section

From Voice Menu, you can access the Voice Recording section.

The list of the items displayed contains all the recording files available on the system:

 

Each recorded call shows the following information:

  • Type: type of the calls (inbound, outbound, internal or dialer)

  • UniqueID: uniqueID of the call

  • Caller: caller phone number

  • Called: called phone number

  • Connected: the internal number of the agent that has managed the call

  • Queue: name of the queue the call was managed

  • Agent: the agent who has managed the call

  • Rating: the rating assigned to the call. You can just click on the recording and choose a value (from 1 to 5 stars). Moreover it’s possible to insert a comment note, visible in the popup that opens when clicking on the registration and also as tooltip by passing the cursor over the rating

     

  • Audio:  click on play button to hear a preview (ONLY for .wav format)

  • Date: date of the recording

  • Disposition / Second Level Disposition/ Third Level Disposition: the labels used if the call has been disposed

  • Transcribe: the status of the transcribe action

  • Sentiment Analysis: the status of the sentiment action 

  • Post Call Analytics: the status of the sentiment action 

In the header you can find two Search fields:

  • search by Date Range

  • search by Value → The tooltip icon links to this wiki to know fields on which filter is applied. In fact search is done on all these columns:

Id

UniqueId

CallerIdNum

MemberName

CallerIdName

AccountCode

UserId

Channel

ConnectedLineNum

ConnectedLineName

Context

Exten

Value

Type

Rating

Queue

Location

FileUri

FileText

CreatedAt

UpdatedAt

UserDisposition

UserSecondDisposition

UserThirdDisposition

TranscribeName

TranscribeStatus

FailureReason

 

Sentiment

sPositive

sNegative

sNeutral

sMixed

tempSentiment

 

 

The three dots menu at the right end of the line   shows the commands that can be applied to the selected recording:

  • Edit Voice Recordings: see LINK

  • Run Transcribe (AI Tool): when you click on it, the transcribe process starts and the label IN PROGRESS is shown until it is COMPLETED

  • Run Sentiment (AI Tool): when you click on it, the sentiment process starts and the label IN PROGRESS is shown until it is COMPLETED

  • Run Transcribe&Sentiment (AI Tool): this command runs transcribe and sentiment at the same time. When you click on it, the process starts and the label IN PROGRESS is shown until it is COMPLETED

  • Run Post call analytics (AI Tool): when you click on it, the post call analytics process starts and the label IN PROGRESS is shown until it is COMPLETED

  • Download Voice Recording: click here if you want to download the selected recording file

  • Delete Voice Recording: click to delete the selected recording file

 

If you want to select multiple recordings and apply commands to all of them in a while, just click the box you find on the left to select the item or click on the box on top to select all of them.

When you have done a selection (of one, more than one or all the recordings), some new commands will appear on the right side of the header:

 

With these commands you can:

  1. Download the details of the recordings (csv format)

  2. Run Transcribe only 

  3. Run Sentiment Analysis only

  4. Run transcribe&sentiment analysis 

  5. Run post call Analytics

  6. Delete the recordings

Voice recordings and permissions

Based on permissions set in the User Profile, Users can see Voice Recordings on these conditions:

  • if the user has permission only on Voice Recordings, he can view the Voice Recordings section with an empty table

  • if the user has permissions on Voice Recordings, Voice Queues and Queue Campaigns, he can view Voice Recordings belonging to assigned voice queues or queue campaigns (internal, inbound, dialer related to queues). So if the user has view of the queues but not the agents, he still views the recordings of the agents assigned to those queues

  • if the user has permissions on Voice Recordings, Voice Queues, Queue Campaigns and Agents, he can view Voice Recordings belonging to assigned voice queues or queue campaigns and recordings of assigned agents (internal, inbound, dialer, outbound)

  • if the user has permissions on Voice Recordings and Agents, he can view Voice Recordings belonging to assigned agents (internal, inbound, dialer, outbound). So if the user has a view of the agents but not of the queues, he still views the recordings of the agents even though they were made on the queues

Edit Voice Recordings

With the top-right commands, you can:

  1. Play the recording 

  2. Download the audio (full voice recording)

  3. Run Transcribe only 

  4. Run Sentiment Analysis only

  5. Run transcribe&sentiment analysis 

  6. Run Post-Call Analytics

 

From the first tab, you can access the General Settings.

Voice Recordings with Quality Analysis

Transcribe, Sentiment Analysis and Post call Analytics are XCALLY Artificial Intelligence Features

You can apply Quality Analysis on New Client Experience: explore here the relative documentation.
Moreover it’s possible to configure AWS Categories find out more

Requirements

How-to change directory for Voice Recordings

Follow below instructions if you want to move the default directory for Voice Recordings store.

  1. Rename the default directory /var/spool/asterisk/monitor:

    mv /var/spool/asterisk/monitor /var/spool/asterisk/monitor_backup

  2. Place a symbolic link. For example, if you you want to move recordings to /usr/src/myrecordings:

    cd /var/spool/asterisk ln -s /usr/src/myrecordings monitor

    VERY IMPORTANT: the name of symbolic link MUST BE "monitor".

  3. If you want to copy the already existing recordings, execute:

    mv -f /var/spool/asterisk/monitor_backup/* /usr/src/myrecordings

How-to change encryption algorithm for Voice files

It is possible to change the encryption algorithm used by the system to encrypt Voice Recording files.

Currently, only one subset of AES flavors is supported. Default value: aes192.

  1. Connect to the server using SSH and edit the hidden configuration file /var/opt/motion2/.env

  2. Update the encryption algorithm, by adding this line, with the new value

  3. Restart XCALLY Motion following this procedure:

Related topics