Calls

In this section, you will find everything regarding the Calls channel inside Conversations and the options we currently offer within it. To enable Calls in Conversations, contact your dedicated Account Manager. Now, let's learn how Calls can help you run your contact center more efficiently.

Technical requirements

Below you can find all the technical requirements you need to know before you start using the Conversations Call Center solution.

System requirements

Audio and video call quality can be very low and unsatisfactory if agents do not have the necessary computing power. Therefore, for a smooth experience, we recommend that the agents' computer has at least the following specifications:

• 64-bit operating system
• 4 GB RAM minimum (8 GB RAM recommended)
• Dual-core processor

Operating systems

Conversations is a cloud-based application that runs on your browser so any desktop-operating system that can install a supported web browser, such as Windows, MacOS, Linux, and Chrome OS is needed.

Peripheral devices

To make or receive calls, agents need the following equipment:

• Headphones or speakers
• Microphone
• A camera that your browser can recognize as a video device (for Video calls)

Additional requirements

JavaScript has to be enabled. Also, you need to grant permission to use the microphone and camera (for Video calls) for Conversations.

Media stream requirements

It is worth mentioning that media streams are mostly symmetric. What you send is about as much as you can receive. There can be exceptions when the link quality is asymmetric.

Initial Conversations web page download is about 11MB, so the initial load can take a long time if there is no adequate bandwidth. The recommended bandwidth for agents to use the Conversations web application uninterruptedly is 5Mbps.

Additional bandwidth requirements per media stream type can be found below:

Media Type	Details	Required Bandwidth
Standard Audio Quality	Narrowband to wideband	50Kbps
High Audio Quality	Up to fullband, stereo	100Kbps
Low-resolution Video	240x180	180 Kbps
SD Video	640x480	600 Kbps
HD Video	1280x720	1500+ Kbps

Media link quality

It is also quite important that you have a solid internet connection for optimal media experience, considering latency and packet loss can negatively impact call quality.

• Packet loss is under 1%
• One-way latency < 150 ms

Network topology variants

When calculating internet speed requirements, one needs to take the following into account:

1. Where are the agents located?
2. Do you have BYOC (Bring Your Own Carrier) or are you using Infobip voice connection?

The table below covers the audio portion only and assumes that offices use the same internet link for BYOC, SIP trunk to Infobip, and SIP trunk to telecom:

Agent Location	Trunk	Required Office Bandwidth per Call	Agent Home Bandwidth
Home	IB	N/A	5Mbit for conversations, after that ~50kbps for audio.
Office	IB	5Mbit for conversations, after that ~50kbps for audio.	N/A
Office	BYOC	2x100Kbps for BYOC g711 audio. Mandatory < 1% packet loss.	5Mbit for conversations, after that ~50kbps for audio.
Office	BYOC	2x100Kbps for BYOC g711 audio. Mandatory < >1% packet loss. 5Mbit for conversations, after that ~50kbps for agent audio.	N/A

Data centers

Calls in Conversations are relayed to Infobip data centers. To minimize latency, use the nearest data center that matches your agents' physical location.

DC	Location	Info
Frankfurt	EU	Full solution
Mumbai	India	Full solution
New York	US	Full solution
Sao Paolo	Brazil	Media only

Codecs in use

Audio	Video
OPUS, PCM	VP8, H.264

Browser compatibility (web)

We support up to 5 most recent versions (unless otherwise indicated) of the most commonly used browsers:

	Google Chrome	Mozilla Firefox	Safari*	Microsoft Edge**	Opera	Internet Explorer
Linux	✔	✔	N/A	N/A	✔	N/A
MacOS	✔	✔	✔	✔	✔	N/A
Windows	✔	✔	N/A	✔	✔	N/A

*WebRTC support in Safari started with version 11.
**WebRTC support in Microsoft Edge started with Chromium-based version 79. Only Chromium-based Microsoft Edge versions for Windows are supported for InfobipRTC.

Browser compatibility (mobile)

For mobile, we support the following browsers for WebRTC calls via call link:

	Google Chrome	Mozilla Firefox	Safari*	Microsoft Edge**	Opera	Internet Explorer
Android	✔	✔	N/A	✔	✔	N/A
iOS***	✔	✔	✔	✔	✔	N/A

*** WebRTC support in browsers other than Safari started with iOS version 14.3.

Call features overview

Call features	Phone calls	WebRTC calls
Queues	✔	✔
Routing	✔	✔
Automatic Agent Assignment	✔	✔
Working Hours	✔	✔
Waiting Strategy	✔	✔
Recording	✔	✔
Make a Call	✔	✔
Receive a Call	✔	✔
Call Wrap-up	✔	✔
Video	N/A	✔
Screen Share	N/A	✔
Dialpad DTMF Inputs	✔	N/A
Transfer calls	✔	✔
Hold	✔	✔

Getting started

To start using the Calls channel properly, you need to fulfill a few prerequisites:

1. Create a user (opens in a new tab).

2. Assign Conversations roles.

3. Configure your call center:

   • Configure a Voice number (follow the guide under Set up calls) for Conversations AND/OR
   • Set up WebRTC application for Conversations AND/OR
   • Configure Call link (opens in a new tab).

   To be able to receive calls, agents:

   • Must be in one of the statuses from the available status group.
   • Must be assigned to the queue that the call is routed to. Note that by default, calls are routed to the Default queue.
   • The queue that the call is routed to needs to have automatic agent assignment turned on. Calls cannot be manually assigned. By default, automatic agent assignment is turned off for all queues, including the Default queue.

Microphone and camera permissions

When an agent initiates or receives the call for the first time, they will be prompted (by a browser pop-up notification) to allow Conversations access to the device’s microphone and/or camera. If the pop-ups are blocked, you will have to manually enable these settings.

Also, if you have denied the permissions, the call will be rejected showing the following error: Permission to use camera and/or microphone denied.

Device settings

In their everyday work and while handling calls, agents need to be able to configure the desired input and output devices (camera, microphone, and speakers) quickly and effortlessly for both voice and video calls. There are two ways they can do this:

The device configuration that agents create on the settings is sticky and will be preserved in the browser. Agents can choose one of the recognized devices from the drop-down lists for both audio and video input devices.

Here, they can confirm that the devices are functioning properly:

• Click Test speakers to test the sound. The system will play music for 5 seconds.
• Click Test microphone to test the microphone. The system will wait for the sound and animate the volume bar accordingly, showing that the microphone is working fine.

The video stream from the selected camera will be automatically shown in the preview window.

If available, the devices that the agent selects will be used for the new calls that they establish with their customers.

Agent can also change the input and output devices while on the call - select the Device Settings option in the side panel of the call modal.

To change the current devices, select your preferred microphone/speakers/camera from the drop-down menus of available devices. Changes take effect immediately. Check for the animated volume bars that your devices are working properly.

Numbers

Voice numbers are phone numbers with the capability to receive incoming calls. They can also be used as the Caller ID for outgoing calls.

To set up a Voice communication and receive incoming calls for your business needs, first, you must lease a Voice number from Infobip.

Refer to Voice Number (opens in a new tab) for detailed instructions.

Bring Your Own Carrier

Bring Your Own Carrier or BYOC is a feature that allows our customers to use our platform, regardless of who is the owner of the DID number and connection towards the operator.

This solution provides flexibility by defining a SIP Trunk (opens in a new tab) between existing customer connection - voice provider and Infobip platform.

Depending on the use case and client requirements we can provide both inbound and outbound configurations. For any clients that require enhanced security, we can set up additional encryption - encryption per call or encryption of tunnels between two data centers.

Contact your Account Manager for any inquiries regarding the BYOC setup.

System configuration

The following section provides everything you need to know about the Calls setup inside Conversations and how agents can receive and make calls, as well as the details about video call options available to agents during their conversations with customers.

For additional details on how to setup the general behavior of a contact center, such as for queues, routes, SLA, etc., see the Supervisor guide.

Generate Call link

The Call Link feature enables efficient support without the need for physical presence in the same location. With just a few clicks, you can engage in video or audio conversations with your customers, eliminating the necessity for any web or mobile application development efforts.

Agents can use the Call link feature to create and send web links that your customers can access and initiate a call to your Conversations contact center. This offers the flexibility of an audio or video call based on your customer's preferences and setup.

If you have not created any setup, the default call link configuration will be used.

To set up and modify the behavior associated with the Call link according to your preferences, it is necessary to:

1. Navigate to Channels and Numbers → WebRTC → Call Link → Configurations, and create a new configuration according to your preferences.

   1. More information on how to create the configuration can be found here.

2. Go to Conversations → Settings → Call Link and apply the created configuration in the Default setting for Conversations drop-down menu.

3. Change the desired Routing options:

   1. Route to agent who shared the link.
   2. Use existing routing rules.

4. Choose what Link type you want to use:

   1. One-time link (individual customers can use this link for single access).
   2. Time-frame link (customers can use this link within a specified date and time range the agents set).
   3. Expiration link (customers can access this link multiple times, starting from its creation until it reaches its expiration time you set up).

Customer experience

When the customer receives a generated call link and clicks it, the web page opens containing the following options:

• Start audio call
• Start video call

During the call, customers have the option to:

• Drop the call
• Mute
• Share screen
• Turn on/off camera
• Switch between back/front camera (available only on devices that have multiple cameras such as smartphones, and tablets)

Wrap-up window setup

To define the default time to be used for this activity, Supervisors need to navigate to Conversations → Settings → General setup.

Click the Call Wrap-Up and set the value.

Adjust call wrap-up

The wrap-up time can be altered at the agent level because not all agents manage the same call types and do not have the same skills. To make this change, navigate to Conversations → Settings → General setup.

Click the Agents tab and select the agent. Then adjust the values based on your needs in the Call Wrap-Up tab.

Recording

Whether it is to adhere to regulatory/compliance requirements, ensure quality, increase operational efficiency, improve the customer experience, or any other reason specific to your business needs, you may need to record conversations between agents and customers.

With Conversations, you can record the following types of calls:

• Audio calls
• Video calls
• Call Link (audio/video) calls

Every call inside Conversations has multiple participants. A participant can be a customer, an agent, a chatbot, etc. Each participant is considered a separate leg of the call. Each participant's leg is recorded and stored as a separate audio file.

For example, a conversation between a single customer and an agent will have two different recording files since there are two legs, or participants, in the call. Conversely, a call that starts between Customer A and Agent 1, which is ultimately transferred to Agent 2, will result in four separate audio files, two for the conversation segment between Agent 1 and Customer A and two for the conversation between Customer 1 and Agent 2.

Conversations allow you to choose from a flexible set of configuration options. You can choose to have a single recording configuration for the entire organization or create granular recording compositions for each agent.

To configure and enable recording, navigate to Conversations → Settings → Call add-ons → and click the Call recording tab.

You can modify configurations only when recordings are disabled, i.e., turned off. To change settings, make sure that you set the toggle at the top right of the recording configuration page to Disabled.

You can choose between 3 different options for your recording needs:

• Inbound and outbound
• Inbound
• Outbound

Conversations also support video calls. Use the Video recording options section to alter this default setup. You can choose one of the following options to record these types of calls:

• Audio only - Records only voice. Video is not recorded.
• Video and audio - Records video and voice.

Besides this, you can choose between two options for the recorded file composition:

• Save each call leg separately
• Merge both call legs

If you decide to save each call leg separately, it would mean that each call leg would have its recording file, and if you choose to merge both call legs, you would have one merged file for every call (except transferred calls).

Storage

You can choose between the Secure File Transfer Protocol (SFTP) and Cloud solution to store your recordings.

As mentioned previously, you can choose what kind of recording files will be stored on servers - separated by call legs or as one unified file. When you download the recording file from the cloud storage or search for the file in your SFTP folder, you can have it composed as a single file which enables you to listen to the whole file and understand the context of a specific call.

Default Cloud Storage

By default, all recording files are stored encrypted on Amazon Web Services. The Amazon Web Services region where the recording files are stored depends on the location of the Infobip data center, which is home to your account.

The table below shows the most up-to-date location map of Amazon Web Services to Infobip data centers:

Region	Infobip Data Center	Recording Storage Location
Europe	Frankfurt	AWS: eu-central-1
North America	New York	AWS: us-east-1
Brazil	Sao Paolo	AWS: sa-east-1
India	Mumbai	AWS: ap-south-1
Indonesia	Singapore	Locally at Infobip DC
Saudi Arabia Saudi	Saudi Arabia	Locally at Infobip DC
Turkey	Turkey	Locally at Infobip DC

Recording files are segmented by customer account so that no customer can access any other customer’s files. A fair usage policy governs the monthly amount of storage per agent, in gigabytes, available to each customer. The amount of cloud storage depends on the license level you purchase.

The storage total is allocated per customer account. For example, if a customer is given 5GB of storage per agent and the account has 10 licensed agents, the customer is allocated 50GB of storage per month. You can purchase extra storage space per month as needed.

Secure File Transfer Protocol (SFTP)

Around the world, many government-led initiatives, laws, and regulations aim to protect their citizens' data. In many regions, storing recording files out of the country is illegal. Therefore, you have the option to move recording files off of the Conversations platform for local storage. The Secure File Transfer Protocol (SFTP) feature enables you to upload recorded files to a secure location.

Before configuring Conversations as a product, you must have a previously provisioned, functioning, and available SFTP server. If needed, consult with your local IT professionals before configuring SFTP server configurations.

You can access the SFTP configuration in the Call recording section under the Conversations Settings → Call add-ons page. Click the Configure button located at the bottom right.

To successfully configure SFTP, you will need the following information:

• Host IP address or DNS name
• IP Port used by the host
• Storage path where files will be transferred
• Username and password

After you have completed your SFTP server configuration, use the Upload recordings to the SFTP toggle button to enable the SFTP transfer.

Recording file name logic

File recordings are stored in a meaningful way, and when you download one or preview them from your SFTP, it is easy to identify the ones belonging to a certain agent or customer. Additionally, the file name contains the Conversations ID as well as the date of a specific call.

The recording name appears in the following format:

• <conversationID> _ <agentname> _ <end-user> _ <UTCStartTime>
• <UTCStartTime> is displayed as YYYYMMDDHHMMSS (using the 24-hour clock format)

Conversations data inside recordings metadata file

If you are using the SFTP storage option, you will have an additional .JSON file that contains the complete metadata for that recording file. The file is populated with Conversations-specific fields, and when you are searching for a recording file, you can identify it by, e.g., agent name, Conversation ID, etc.

The .JSON file stored on SFTP is populated with Conversations specific data, such as:

• Agent name
• End-user (name, if it exists, and phone number)
• Conversations ID
• Timestamp (UTC Start Time and UTC End Time)
• Conversations ID
• Call ID

When you are using the Cloud storage option and download the recording file, the downloaded .csv file is populated with Conversations specific data as well:

• Agent name
• End-user (name, if it exists, and phone number)
• Conversations ID
• Timestamp (UTC Start Time and UTC End Time)
• Conversations ID
• Call ID

Configure recording for agents

All calls are recorded by default when the top-level recording toggle is enabled. There may be various reasons why you may want to disable recordings for specific agents. For example, if compliance regulations prevent certain types of calls from being recorded, it is necessary to have more granular recording configurations. This is where Conversations allow you to override the default configuration on an agent-by-agent basis.

To set up more granular compositions, navigate to the agent’s profile under Conversations → Settings → General setup. Click the Agents tab. Select the agent you wish to configure and then click the Call recording tab.

As with the top-level recording toggle, you must set the Record Agent toggle to Disabled to change the recording parameters for the agent. Disable agent recording and then make the desired selections under the Recording configuration, selecting either inbound, outbound, or both under the Custom drop-down menu option. Deselecting the checkbox disables recording for the given option.

How to work with recordings

Once your recording setup is complete, calls will be recorded based on your configurations. The Conversations UI provides easy access to playback, download, and deletion of recorded files.

Depending on the configuration you have chosen, each participant's leg can be recorded as a separate or unified file.

In the Conversations UI, you can choose to listen to the full stereo recording (all legs) or each participant file independently (depending on the configuration). Recordings are conveniently located inside each conversation.

In the All Work tab, navigate to any call conversation with a recording file. There you will see a quick link to the stereo playback option. Click the play button to launch the audio player and the playback will begin automatically.

You can pause and resume playback for both audio and video recording. Also, you can fast forward or backward the recording file on both audio and video calls by sliding the progress bar or set a custom start point to listen to a specific recording part.

By hovering over the speaker symbol, you can adjust the volume or mute the sound.

If needed, you can play the video recording in full-screen mode by clicking the respective button.

Finally, you can close the player at any time during playback.

To access the recordings, whether as separated legs or unified files, navigate to the Recordings tab. You will notice a list of individual recording files in the table shown in the UI. Click the Play icon to launch playback for the single file you selected.

You can view additional metadata and details about the recording by clicking the > symbol to the right of the three dots, as shown in the image below.

Download recording files

You can download call recordings from the Conversations UI, and you can do so with each file. There is currently no bulk-download option for recordings. To access downloads, click the three dots and then click the Download recording button. This will download the recording file and a CSV file with associated metadata related to the recorded conversation.

Delete recordings

You can also delete call recordings from the Conversations UI and do so for each file. The significant benefit of separate recording files is that you can selectively delete them. For example, suppose a European Union citizen exercises their right to be forgotten. In that case, you can delete the file for that specific citizen while still preserving the agent side of the conversation (provided there is no PII data in the recording) for use in quality management or customer experience improvement activities.

To delete a file, click the three dots and then Delete.

Although the recording files are removed from the platform once deleted, there is still a record that the recording existed. While the recording itself is no longer available, Conversations maintains a history of the previous existence of the recording along with an audit trail showing who deleted the file and the deletion date for a certain period of time, and this is deleted as well after the arranged retention period.

Data retention and deletion

Infobip deletes the collected data to ensure not to keep it longer than necessary, depending on your business needs and any applicable legal regulations.

Unless otherwise agreed with you, the default retention rules are implemented to take care of the collected recording data. However, if your business needs require retention periods that differ from the default period, you can customize the retention period based on your needs.

You can find out more about data retention on this page.

Custom transfer music and message

Supervisors have the option to configure custom transfer music and message according to their business preferences, resulting in a seamless customer flow. By default, the preselected music is played for all of your transfers.

In order to change this behavior and configure your custom transfer music and message, navigate to Conversations Settings → Call add-ons → Call transfer. Once the page opens, the Transfer music and message modal will display two options to configure:

• Music (plays default music or your uploaded audio file)
• Message (plays a message or an audio file in a defined interval)

Transfer music

In the Transfer music modal, you can choose which music will be played during the transfers to an agent or external number:

• Default music (enabled by default)
• Upload file (custom music)

If you want to upload your audio file, click the Upload file option where you can either click the Browse button or drag a file to the dedicated area and upload it.

Once the file is uploaded, you can listen to it by clicking the Preview button.

After you finish editing, click the Save button to apply the changes.

Transfer message

To set up the transfer message, click the Message option in the aforementioned modal. Whether you decide to set up your transfer message as a speech version of the text you provide in a written format, use the Text-to-Speech option and write the message script in the provided text field. Once you write out the audio message content, configure these available parameters:

• Language (select one of the available languages)
• Voice (select one of the available voices)
• Speech rate (adjust the speed at which words will be spoken to be in line with the language and the message you are communicating)
• Repeat Interval (choose how frequently you want the message to repeat)

However, if you decide to use file upload as the audio source, click the File Upload option to browse for the desired audio file (or use the drag and drop action to upload your file), locate it, and confirm the action to upload your file. Once the file is uploaded, you can also set the Repeat Interval.

When you configure the transfer message, do not forget to enable and save the changes. Now, during the transfers to an agent or external number, the customers will hear the configured music and/or message in specified intervals you have set up.

Call time-out configuration

Call center supervisors can configure the custom behavior of the system and set a custom call time-out for incoming calls (including call transfers). With this option, you can specify for how long should a call ring to one agent before moving to another one.

If the agent ignores the call and the call time-out is reached, the call is moved to another available agent (respecting the routing logic).

To configure the call time-out, go to Conversations → Settings → Call Add-ons → Advanced controls.

Here, you can specify how long the call will ring or reset the value to its default setting (30 seconds).

Waiting strategy

Being on hold can be frustrating for your customers so you must be able to build a waiting strategy that establishes the best environment for the customers while they wait to be connected to your agents.

To do so, navigate to Conversations → Settings → General setup. Click the Waiting Strategies tab, and to define a new waiting strategy, click the Create New button.

Name your waiting strategy and define general information:

• Numbers or application (when using Web and In-App Calls) that your strategy applies to.
• Maximum waiting time that defines for how long this strategy will run.

Subsequently, you can set up one or more available options depending on your business needs:

• Greeting message - This will be the first communication customers hear when they contact you. You can use the text-to-speech feature to create a message or upload a file you would like to play to your customers.
• Hold music - Certain types of music can calm a person while on hold and create a more positive experience for customers. You can upload your hold music file.
• Announcement message - When your customer's mind is occupied with something it makes the time they have to wait feel less burdensome. Keep customers engaged by periodically playing announcement messages.
• Maximum Wait Time Message - When the maximum wait time has been exceeded, and just before the call is about to be dropped, play a message to your customer letting them know that the call will end. You can use the text-to-speech feature to craft a message or upload a file you would like to play to your customers.

You can enable/disable each option by clicking the toggle button.

Create message

You can decide to create your waiting strategy messages as a speech version of the text you provide in a written format. In that case, you should use text-to-speech as an audio source and write the message script in the provided text field.

To do so, configure these available parameters:

• Language - Select one of the available languages.
• Voice - Select one of the available voices.
• Speech Rate - Adjust the speed at which words will be spoken to be in line with the language and the message you are communicating.

Upload file

You can use a file that you have previously recorded or created in your waiting strategy messages. In that case, you should use file upload as the audio source. Click Browse (or use drag and drop action to upload your file), locate the file, and confirm the action to upload your file. Also, you can delete uploaded files if needed.

Automatic agent assignment

You can set up the automatic agent assignment for calls. Go to Conversations → Settings → General setup → Queues.

When a new conversation is assigned to a queue:

• The automatic agent assignment process starts (only one can run at any given time).
• It is applied to newly created conversations.
• Or when an agent changes their status to Available.

For calls, availability is taken into consideration. An agent is considered available to take on new calls when they set their status to Available, and they have fewer conversations assigned in any status other than Closed, than their defined capacity.

Workload and waiting times are calculated for every agent who is in the Available status.

To see more details about the automatic agent assignment configuration, click here.

Callback

The Callback feature allows customers to request a callback from the agent to the phone number associated with the conversation.

Callbacks aim to improve the customer experience by minimizing the frustration of long wait times or when calling a business outside of working hours. A callback can be triggered by a long wait time, a lack of available agents, or if the customer simply wants to be called back later.

Here are the 3 steps to enable callback for your customers:

1. Create the Callback configuration.
2. Apply the Callback configuration.
3. Set the action code to trigger the Callback.

Follow the Agent callback section for more information about the agent's experience with callbacks.

Configure callback

To activate a callback, you first must configure the feature to achieve the desired system behaviors. Navigate to Conversations → Settings → click the Callback tile. Now, click the Add New button.

A new screen opens where you can specify the desired selections for your callback configuration.

At the top of the configuration page, name the callback configuration so it is easy to identify it on later screens.

Next, choose the desired callback type:

1. Automated

   • This option ensures that the callback process is completely automated. Using the existing routing and prioritization rules, the mConversations platform will select and assign an agent to the callback.
   • The callback will automatically ring on the agent’s desktop or mobile app.
   • When the agent answers, an outbound call is made to the customer’s callback number.
   • If the customer does not answer, the call is assigned to the next available agent.

2. Manual

   • The Conversations platform maintains a list of callbacks.
   • Agents need to navigate to All Work → Callbacks and manually select the callback request to handle.

Additionally, the priority of the callback can be set based on one of two options. These are only available if the callback type is set to Automated:

• Inbound priority
  • Prioritizes incoming calls (direct or in a queue) over callback requests.
  • After all incoming calls have been handled, callbacks are assigned to available agents.
• Queue position priority
  • Calls are prioritized based on their place in the queue. The Conversations platform remembers the customer’s position in the queue and assigns them to agents in the requested order.
  • New incoming calls are considered to have lower priority.

You can configure the number of callback attempts. The configuration is universal across all callbacks regardless of where they are received. Callback retries to ensure that multiple attempts are made to contact a customer who does not answer the call when the agent initiates the callback.

There are two parameters available to configure:

1. Number of callbacks:

   • Determines the number of times a callback will be attempted in the event a customer does not answer. Note that the first callback attempt is not considered a retry, and thus is not included in the number. For example, to ensure that 5 total attempts are made to contact the customer, you should enter the value of 4 in this field. This value ensures that after the first attempt fails, exactly 4 more attempts for a total of 5 attempts will be made. When the maximum number of attempts is reached, the callback is removed from the queue and closed.

2. Delay between callbacks:

   • The time value in this field indicates the delay between unsuccessful retry attempts.

You can also configure a configuration message, which is played after the customer enters a response code that triggers the creation of a callback request. These messages can be either text-to-speech or pre-recorded audio file messages. The confirmation message is globally applied across all callback types and triggers for the given callback configuration.

Text-to-Speech

Use this feature to craft a custom message. If you select text-to-speech, specify the following items:

• Language - Select one of the available languages
• Voice - Select one of the available voices
• Speech rate - Adjust the speed at which words will be spoken to be in line with the language and the message you are communicating

Upload Audio File

Drag and drop or browse to add a pre-recorded audio file.

Apply configuration and set action code

For callback requests to be functional, you must link the callback configuration to a specific Announcement message under the Waiting strategy and the Away message under the Working hours sections.

This is how you can link the two:

1. Create a custom text-to-speech message, or upload a custom audio file.
2. Set the response code to trigger the callback.
3. Link the response code to the callback configuration created in the previous section.
4. Set the time to wait for a response from the customer.

To enable Callback for calls in queues, set up the announcement message for the Waiting strategy:

• Navigate to Conversations → Settings → General setup → choose the Waiting strategies tab and open an existing strategy → click the Announcement message tab
• Alternatively, click the Waiting strategies quick link located on the Callback configuration page and then click Announcement message.

To enable Callback for calls received after business hours, set up the announcement message for the Waiting strategy:

• Navigate to Conversations → Settings → General setup → choose the Working hours tab and open existing configuration → click the Away message tab.
• Alternatively, click the Working hours quick link from the Callback configuration page.

To set up Response codes, scroll down this page, click the Add New button, and fill in the following:

• Number: Add the desired DTMF code to be applied to the Callback.
• Action: Choose Callback as a value from the drop-down list.
• Value: Select the callback configuration you want to apply.

You can configure the waiting time for response/input from your customers in Wait for response codes field.

Callback requests list

To preview the list of callback requests, go to the Conversations tile on the left menu → choose All Work → click Callback under Global Views.

There you can see all callback requests with predefined filters:

• Callback status = In queue
• Callback type = Manual
• Conversation status = Open, Waiting, or Solved

Callback-related fields and their corresponding values are as follows:

• Callback Type: Manual, Automatic
• Callback Status: In Queue, Active, No reply, Done
• Callback Pending Since: Predefined list with a time period
• Callback Starting Date: Predefined list with a starting date, calendar input for start and end date

Call on hold

To configure music on hold and the announcement message that the customers will hear while waiting for the agent to come back to the call, navigate to Conversations → Settings → and click the Call add-ons tile.

Here, you can configure the Default message by clicking the three-dot menu (or the straight line) and choosing Edit.

If you want to create a new configuration, click the Create New button.

Choose music that will play while customers are on hold, or upload your audio file. Before you save the setup, you can preview the music that will play. Add the content for the announcement message or upload an audio file.

For the announcement message, make sure to enable the message to be played out.

Then, choose whether this is going to be text-to-speech or your file that you can upload and play as the announcement.

If it is text-to-speech, add the desired announcement message. Then select the language, the voice, the speech rate, and how often you want the message to be repeated (default is every 10 seconds).

Hit the Create button to save the configuration for this queue. You can do the same for any other queue if you wish to have a different setup.

Transfer to external number

Agents can use the Transfer to external number option to transfer customers to any valid E.164 formatted phone number. Phone numbers can be stored in People under Contacts or entered manually by an agent.

An organization needs to add the external_number tag to the profile of any contact (existing or new entry) in People they want to have available for external dialing. People allows you to build rich customer profiles by unifying customer information across the Infobip product suite and your disparate back-end systems.

To configure the Transfer to external number feature, navigate to Conversations → Settings → and click the Call add-ons tile.

Once the page opens, click the Call transfer tab where you can see a toggle for whether to allow an agent to manually enter phone numbers that are not stored in People when transferring a call.

If the option is enabled, the agent can transfer the call to any number even if it is not added to People. After the call is transferred to this number, the system will add it automatically to People and the number will appear in the drop-down list the next time.

If the option is disabled, the agent cannot transfer the call to a number that is not stored in People.

When the option Numbers not stored in People is enabled, agents can manually enter transfer destination phone numbers in the portal UI. The entered number is automatically added to People with the external_number tag and the next time, the agent will have it listed in the external number phone book and it will be suggested as the agent begins to type it.

As soon as you set up the Transfer to external number functionality, agents will either be limited to external phone numbers stored in People and/or can enter phone numbers manually. The flow varies slightly depending on the settings:

• While the agent is connected to an active call, the transfer options become visible to them where they first need to click the Transfer button.

• The From field is the number shown to the external party during the transfer (if the customer's initial call is PSTN, the From number drop-down menu is not offered because the system automatically uses and assigns the customer's DID number as a From number; if it is a WebRTC call, the drop-down menu is shown since the customer's DID number is unknown, and the agent has the option to choose one of the DID numbers already assigned to the account).

  • They can choose a number from the list menu or leave it for the system to choose automatically.

• The To number field represents the destination phone number for the external transfer where the agent can enter a name or a phone number based on the following:

  • If the application is configured to allow phone numbers to be entered manually, the agent can enter the desired destination phone number
  • A number is offered if a person in People has a phone number to be shown
  • A person from the external number phone book is shown only if they have the external_number tag assigned to them
  • If they start typing a name or a phone number for a specific person/third party, all phone numbers for a People contact are shown

• The agent needs to click the Transfer button.

• Once they transfer the call, they are disconnected from it and the external party continues the conversation with the customer.

Agent guide

Everything you need to know about the agent's capabilities and options, including those for receiving calls, making calls to customers, conducting video calls, etc., is covered in this section.

For additional details about the agent behavior, such as how to log in, change status, create a note, etc., see Agent guide.

Receive a call

Calls can be placed to phone numbers or they might originate from an application. It can be either audio or video. When a call is received, a new conversation is created, or the call is included in an existing conversation.

You can receive only one call at a time. To receive the call, you must be in the Active status as well as in the queue that the call (conversation) is routed to. Agent is considered unavailable while in wrap-up time. Read more on wrap-up under the Wrap-up window section.

When a call is routed to you, you will hear a ringing sound and an incoming call pop-up window with the customer's name will be shown in the Agent panel. This will happen if the search based on the phone number in People is successful. Otherwise, only the phone number will be presented to you.

The pop-up window will also enable you to answer or decline the call.

Incoming call information

The incoming call modal is the very first information the agents may find out about the customers. We have enriched it to showcase more information about the customers and the incoming call.

Before answering the call, agents can now see this info on the call modal:

• The name of the queue the call is coming to (e.g., Customer support).
• Whether this is a regular call, a transferred call, or a callback call.
• Customer's full name and phone number.

Moreover, agents now have the option to copy the phone number with a single click and paste it into their notes, send it to a colleague, or a third-party CRM if used for any storage information purposes.

Make a call

Sometimes agents need to call the customer first. Conversations offer the option of making a call to your customer's phone number.

To do that, click + inside the My Work panel to start a new conversation and specify the following:

• In the From field select one of the available Voice numbers. For the number to be shown in the list it needs to be configured for Conversations. Read more on how to configure numbers (opens in a new tab).
• In the To field you need to specify the contact you will call. You can do that by searching and selecting some of the contacts from People or by typing in a new number and adding it.

Click here to find out more about the phone number validation.

Then click the Start Conversation button to create a new conversation thread.

The system will validate if there is an ongoing conversation and if there is, will offer the following scenarios:

• Pre-existing conversations assigned to the logged-in agent or unassigned
  • The agent is prompted to view the conversation and continue the communication with the customer if it is relevant to the ongoing topic. A follow-up call can be initiated if needed.
• Pre-existing conversations assigned to another agent
  • The agent is notified that a new call cannot be initiated, nor a conversation started with this customer as another agent is already communicating with them. In this case, both parties need to agree on whether the conversation can be transferred; another agent can make that call or the ongoing conversation be closed.
• No existing conversations
  • A new conversation will be created

A call from the agent towards the customer is initiated and a call pop-up window is shown.

When making outbound calls towards a phone number, agents will hear certain pre-connection audio (Early Media (opens in a new tab)) such as busy signals, failure-to-connect errors, or other informational messages from mobile operators.

After the customer has answered the phone, a new pop-up with call controls will be displayed.

Make an outbound call

When creating a new outbound call, you can add additional fields to the conversation, such as:

• Form (optional)
• Priority
• Tags (optional)
• Queue

This enhances the level of control over call routing and organization. Additionally, it ensures that sensitive call details are accessible only to authorized personnel.

When creating outbound calls, you can only add a queue you are part of, and you cannot change the assignee.

Make a follow-up call

After you have been contacted by the customer, you can also call them back. Change the channel to Voice, select the number and use the Start Call button shown in the reply message box.

Create and send call link

Agents can use the Call link feature as a fast and simple way to engage your customers in an audio or video call. They can create and send branded customized web links that your customers can access and use to initiate audio or video calls towards the Conversations contact center.

Links can be created for existing conversations as well as new ones. If calls are to be scheduled, time-frame input from agents are required.

Create call link for guest conversations

Agents can create call links unrelated to existing conversations. To do that, agents need to click on the 3-dots menu in the Conversations window and choose the Generate call link option.

If the link type has been configured as time-frame, then agents will be prompted to enter the customer's identity as well as the link validity period in a pop-up window.

Customer's identity will be shown as a caller to the agent on the incoming call. If their phone number is used as identity, the call will be matched to customer's data in People which is when name and last name derived from People will be displayed.

In cases with the one-time link or expiry link, a pop-up window will prompt the agent to enter only the customer's identity.

In both cases if you click on the Generate button, a link for the call will be created. The link can be copied to clipboard and shared with the customer over any available channel.

Create call link from existing conversation

If an agent is already engaged with the customer in an existing conversation via one of the available channels and they want to create and share the call link, agents can use the Generate call link icon in the message box.

If the link has been configured as time-frame, then the agent will be prompted to enter the link validity period in a pop-up window. Clicking on Generate will create the link.

If the link type has been configured as one-time or expiry, the link will be generated automatically upon clicking the Generate button.

Agents can send the link immediately or copy it and send it as part of the next message intended for their customer.

After the customer has received the link, they will be able to access the web page in line with steps explained in the Call link configuration and setup. When they make a call using this link, agent will receive the call in the Conversations panel.

Call options

Here you will find all audio and video call options that agents need both during a call and after the call ends with the customer.

Call controls are dynamic and shown when applicable. For example, video and screen sharing are not available during a regular phone call. Call options can vary, but the most common are the following:

• End call
• Mute/unmute
• Turn camera on/off
• Share screen
• Transfer
• Hold
• Settings

Write messages during a call

During an active call, agents can simultaneously write messages or emails to customers, enabling them to perform various operations at the same time and enhance the overall customer experience. This option is also available to customers who can write and receive messages from an agent.

Agents can initiate a new conversation such as an email or live chat during a call. However, starting a new call is not possible as an agent can only engage in conversation with one person at a time.

If the customer is the same person the agent is currently speaking with, no separate conversation will be created; instead, the messages will be incorporated into the conversation related to the ongoing call.

If the customer differs from the one engaged in the ongoing call, a new conversation will be created.

Screen sharing

Agents can use screen sharing during WebRTC calls. Screen sharing allows them to share their computer screen. More specifically, they can share the application window or a browser tab with the customer who is connected to the call. Depending on the browser agents are using, they may not be able to share some windows/screens.

Background options for video calls

The agents have various functionalities to customize their background during WebRTC video calls:

• Blur effect → Ability to blur the background.
• Face tracking → The camera always prioritizes the user's face, keeping it in focus.
• Custom backgrounds → Users can upload a personalized image to serve as the background during the call.

To customize your background during a video call, click Settings → Video effects.

The menu offers three options (the default background effect state is None):

• Blur
• Face trace
• Add new

If you choose the Blur option, once a video call starts, your background will be blurred by making the area behind you indistinct. This helps bring focus to you while minimizing distractions from what is behind you and providing privacy.

The Face trace option identifies and tracks your face within a video by following the movements and positioning of your face.

The last option, Add new, offers the addition of a customized background to be used during video calls. This feature enables incorporating a personalized background of your choice during video calls.

Transfer to agent

At this time, we only support blind (cold) transfer, which means that the call is not announced to the agent before being transferred.

The other agent you selected (must be from the same queue the conversation landed in) to receive the transferred call will ring and can either accept or reject the call.

To transfer the call, click the Transfer button, then select the Agents in '... queue' tab (the name is dynamic as it depends on how the queue is called). Then select or search for the available agent from the current queue you want to transfer the call to.

What is also important to note is that available agents are shown first on the list, and you can also use the refresh button next to the list to populate it in real time.

When you transfer the call, it will end on your side and continue with the transferred agent and the customer. Additionally, the customer hears the ringing sound while being connected to the other agent. If you click the Hang up button, it will end the call for both parties.

Transfer to queue

You can also transfer calls to another agent in the current queue, or to another queue. Note that you cannot choose a specific agent from a different queue. To do that, while on a call, click the Transfer button and simply choose the desired queue in the call modal drop-down. Alternatively, choose an agent from the same queue.

The drop-down options during transfer show which queues and agents are available in the order of their availability. Statuses are shown in real time - click the refresh button to update them. In case the transfer is unsuccessful (some error occurred, agent suddenly went offline or busy), the call modal will reflect this by showing a warning that the call cannot be transferred.

Transfer to external number

Agents can use the Transfer to external number option to transfer customers to any valid E.164 formatted phone number. Phone numbers can be stored in People under Contacts or entered manually by an agent.

A system administrator or supervisor needs to pre-configure the application to enable external transfers. Based on the settings, agents will either be limited to external phone numbers stored in People and/or can enter phone numbers manually. The flow varies slightly depending on the settings:

1. While connected to an active call, the transfer options become visible to the agent where they first need to click the Transfer button.
2. Click the Number tab to make the transfer to an external number.
3. The From field is the number shown to the external party during the transfer and this field allows you to choose a number from the list menu or leave it for the system to choose automatically.
4. The To number field represents the destination phone number for the external transfer where you can enter a name or a phone number based on the following:
   • If the application is configured to allow phone numbers to be entered manually, enter the desired destination phone number.
   • A number is offered if a person in People has a phone number to be shown.
   • If you start typing a name or a phone number for a specific person/third party, all phone numbers for a People contact are shown.
5. Click the Transfer button.
6. Once you transfer the call, you are disconnected from it and the external party continues the conversation with the customer.

If the option to transfer the call to numbers that are not stored in People is disabled, and you enter a valid number in the To field, the transfer button is disabled with a message that the number or name is not found in the phone book.

Once you transfer the call, the wrap-up modal shows up allowing you to handle the conversation as you want (e.g., close it) and an event is shown in the conversation indicating that you transferred the call to a specific number or person.

If the call transfer fails due to some technical difficulties, the call is returned to the agent.

Call on hold

When the agent answers a call, among different options, they can decide to put the customer on hold if they need to check something while on the call. The agent will remain in the On hold status until they resume the call.

The agent panel reflects this change and adds a new event Call on hold so, if needed, supervisors can check how long calls stay on hold, whether the issues are resolved, and act accordingly to provide the best customer service.

While on hold, the customer will hear an audio cue (either music on hold or a pre-recorded announcement, depending on the setup) until the agent returns to the call. Supervisors can configure music on hold and the announcement per queue (opens in a new tab) (every queue can have a different setup). The Hold option is available on mobile as well.

Dial pad

When making outbound phone calls, under the three dots menu, agents can find the Dial pad feature. With a dial pad, agents can collect and send DTMF inputs (opens in a new tab) when required to interact with an IVR.

Callback

The Callback feature allows customers to request a callback from the agents.

There are two types of callback that supervisors can configure:

1. Automated

   • Callback requests are assigned automatically to agents.
   • Agents do not have to check any lists nor take any additional actions to locate callbacks.
   • Callbacks ring on the agent’s desktop or mobile device. Agents only have to answer the callback when it is presented to them.
   • The customer number is automatically dialed and connected to the agent.
   

   In case the customer does not answer on the first callback, the Conversations platform will try to call them back as many times as specified by the supervisor before removing the call from the callback queue. The callback retry is automatically logged in the system for execution and requires no additional intervention by the agent.

2. Manual

   • Agents navigate to the callback list manually and pick up a callback request.
   • Agents must manually call the customer independently. Each agent will have an assigned list of callback requests.

   Find out more about Callbacks (opens in a new tab).

   The same list will show up for agents in All Work, where in case of a manual callback they can assign or reassign the conversation to themselves.

   

After clicking the Assign To Me button, that conversation will appear in My Work, where you can preview all related information and initialize a call. Conversations containing a small arrow icon next to them indicate that it is callback.

Once you initialize the call and have finished talking to the customer, you can wrap up the call and close the conversation. The callback status will be changed to Doneafter the customer has answered the call.

Callback statuses and event logs

Regardless of the conversation status, a callback has its own status, and a callback call can specifically go through the following:

• In a queue – The customer requested a callback, the conversation is in the queue list, and an agent is not assigned.
• Assigned – An agent is assigned to a callback conversation.
• Active – The agent is assigned to a callback conversation; the call is currently in progress.
• No reply – The agent called the customer, but they did not answer.
• Done – The agent had a successful call with the customer.

Event logs are part of the Conversations timeline, where you can see the precise time the callback was requested, whether another callback has been asked in the meantime, and if the customer has not responded.

Wrap-up window

When a call is finished, the wrap-up window will show where agents can write notes, call details or a summary of the call before closing the conversation. Besides this, they can rate the call quality.

The wrap-up window lasts for as long as the Supervisor has set the timer. However, if you need more time to write the summary, simply click on the +10 Seconds button to add 10 seconds to your wrap-up time. You can write the notes or the summary of your call on the right side of the agent panel. You can also review this summary later on, if needed.

While the wrap-up window is active, the wrap-up time will be running. Default wrap-up time is 30 seconds and it can be adjusted on system or agent level. While in the wrap-up, the agent is considered busy. The wrap-up time ends when the window is closed. If 30 seconds is not enough to complete your notes, you can click on the +More Time button, adding 10 seconds to your wrap-up time.

Before you close the wrap-up window, you can also set the conversation status. If the conversation is not marked as Closed, any additional communication initiated by the same person will be routed and appended to this conversation.

Recording

The agents have the option to listen to the call recordings (audio and video) in case they need to remind themselves of the call content to note it down and/or update the customer data properly.

You can access the recording files directly in the conversation or by clicking the Recordings tab in My Work.

Once you play the audio or video recording, you can pause and resume playback, as well as fast forward or backward the recording file by sliding the progress bar or set a custom start point to listen to a specific recording part.

By hovering over the speaker symbol, you can adjust the volume or mute the sound.

If needed, you can play the video recording in full-screen mode by clicking the respective button.

Finally, you can close the player at any time during playback.

You only have access to your recordings where you can preview the associated recording information. Besides this, you can download the recording file by clicking the three dots and stream it from an in-built player.

Troubleshooting

In case you run into any issues during or prior to connecting to a call, make sure to read our troubleshooting steps below to fix these issues.

Fix audio and video call issues

If you cannot hear or see your customer, try these steps:

1. Check your hardware:

   • Make sure your microphone, camera, and speakers are plugged into your computer and turned on (make sure that the microphone is not muted).
   • If you connect any of your hardware via USB, check if you need to install software from the device manufacturer.
   • Make sure other programs on your computer are not using your microphone, camera, or speakers.

2. Check if you have enabled microphone and camera.

Enable microphone and camera in Google Chrome

1. To change your browser settings, go to the address bar and click on the lock icon on the left of the website address. This lock icon indicates that information is transferred using a secure server connection.
2. Camera: click the dialog box and choose Allow or Ask.
3. Microphone: click the dialog box and choose Allow or Ask.
4. When finished, click the X in the upper right corner to close the dialog box.
5. Reload the page to Update Settings.

These settings and screens depend on the Chrome version you have installed on your computer. Learn how to manage camera and microphone (opens in a new tab) in the latest version of Chrome.

Enable microphone and camera in Mozilla Firefox

A pop-up similar to the screen above will appear on your computer. Select which camera and microphone you want to use and click Allow and reload the page to access the content.

1. If the camera and microphone have been blocked, click the lock icon to the left of the address bar.
2. Use the microphone: click the X icon to unblock the microphone.
3. Use the camera: click the X icon to unblock the camera.
4. Reload the page, and Firefox will prompt you again if you want to allow these tools.

These settings and screens depend on the Firefox version you have installed on your computer. Learn how to manage camera and microphone (opens in a new tab) in the latest version of Firefox.

1. Restart the call and your computer:
   • Exit and rejoin the call.
   • Restart your computer.
2. Ask your customer to troubleshoot:
   • If you cannot see or hear your customer, ask them to try the steps above.

Error messages

Your connection is weak

When Internet issues are detected, agents get a message indicating this inside the call modal. These issues are visible to the agent and they will see this message: Your connection is weak.

Try these steps:

1. If you are using a Wi-Fi network, move closer to the Wi-Fi router.
2. If the above step does not solve your problem, check your connection status (Wi-Fi signal or internet access). If your signal is not working, turn off your internet router and then back on. Wait for some time and try again.

Customer's connection is weak

If the call is a WebRTC call, then the system can detect if there are any issues on the customer's side. If this happens, the agent will get the following message inside the call modal: Customer’s connection is weak.

Ask them to try these steps:

1. If they are using a Wi-Fi network, ask them to move closer to the Wi-Fi router.
2. Ask them to check their internet connection (Wi-Fi signal or internet access).

Low call quality detection

We monitor average call quality over time and measure packet loss, round trip time, and jitter to approximate a human ranking by an Objective Measurement Method.

Somewhere between 4.3 - 4.5 is considered an excellent quality target. On the low end, call or video quality becomes unacceptable below a MOS of roughly 3.5.

Based on the criteria above, we will indicate internet connection issues when our measures are less than 4.