Voice actions
Last updated
Was this helpful?
Last updated
Was this helpful?
Voice actions enable you to create, forward, answer or hang up a channel call in the automated flow. Additionally, it is also possible to perform actions like create or manage recordings, transcriptions or text-to-speech files. Check below the available voice actions.
You can also find more information in our API documentation for , and .
Flows offers the Create call step, allowing you to create channel calls seamlessly. You can specify the destination using either E.164 phone number format or a flow variable, ensuring compatibility with their desired endpoints.
Additionally, there are Advanced Configurations available, these options can be configured to customise the create call rules to your liking:
Step ID: Identifies the step of the corresponding flow.
Amd Settings: Answering Machine Detection (AMD) allows you to identify whether a call is answered by a human or answering machine, and adjust your call flow accordingly. It is possible to configure the following parameters:
Beep Timeout: If ifMachineNotifyAfter=beep and a machine is detected, we will wait for this amount of time for a beep.
If Machine Notify After: Select between beep or word count.
Speech Locale: Choose the language for machine detection (e.g., en-US).
Speech Timeout: The number of seconds words will be counted.
Word Count: Number of words to take into consideration when assessing machine or human. If actual words > wordCount, it will be considered a machine.
Flow start: Activate flow when call is ringing or call is answered.
From: If you want to override the CallerID. Set the number in E164 format. It must be one phone number of a channel installed in the same workspace.
Max Duration: Maximum duration of the call. Call is terminated after the max duration is reached.
Call Notification / URL: The URL used to publish webhooks for this call. Note: You need to have a webhook subscription created.
Record Start: Start recording the call from ringing or after the call being answered.
Send Keys: This attribute instructs the flow to play DTMF tones upon call connection. It proves beneficial when dialling a phone number along with an extension. After dialling the number, when the flow answers the call, the DTMF tones are sent to connect to the extension.
Ring Timeout: Maximum duration for ringing. Call is considered no-answer if not answered inside the ring timeout.
Workspace Id: The Workspace ID of the call.
Wait For Conditions:
Call was answered: Enable if you want to wait if the call is answered to proceed with the flow.
AMD finished: Enable if you want to wait for AMD results. You can then add a conditional to act differently for machine, human or beep.
Please note that each step within a flow is allocated a Step ID, which aids in identifying the step within its respective flow.
When building a flow, you have the option to implement the Forward Call action on an incoming call. You can either populate the "To" field with the destination phone number you want to forward the call to (in an E.164 format), or with a SIP forwarding address of your choice.
Additionally, there are Advanced Configurations available, these options can be configured to customise the call forwarding rules to your liking:
Amd Settings: Answering Machine Detection (AMD) allows you to identify whether a call is answered by a human or answering machine, and adjust your call flow accordingly. It is possible to configure the following parameters:
Beep Timeout: If ifMachineNotifyAfter=beep and a machine is detected, we will wait for this amount of time for a beep.
If Machine Notify After: Select between beep or word count.
Speech Locale: Choose the language for machine detection (e.g., en-US).
Speech Timeout: The number of seconds words will be counted.
Word Count: Number of words to take into consideration when assessing machine or human. If actual words > wordCount, it will be considered a machine.
Max Duration: Maximum duration for the call. Call is terminated after the max duration is reached.
Ring Timeout: Maximum duration for ringing. Call is considered no-answer if not answered within the ringing timeout.
HangupAfterBridge: Disable this setting if you want the parent call to be continued after the forward call is terminated.This setting allows you to continue the flow after the forwarded call hangup.
From: Use this setting to override the CallerID. Set the number in E164 format. It must be the phone number of a channel installed in the same workspace.
Record Start: Start recording the call from ringing or after the call being answered.
CallNotification/URL: The URL used to publish webhooks for this call. Note: You need to have a webhook subscription created.
Ringtone: Force a ringtone. With this option you select the country ringtone you want to play to the caller. Note: this overrides the network ringtone and early announcements.
Call ID: Identifies the call that will be connected to the destinations
Voice Channel ID: The ChannelID that will be used to place the call. Defines the CallerID of the call.
Workspace ID: The Workspace ID of the call.
Wait For Conditions: Enable if you want to wait if the call is answered to proceed with the flow.
Get channel call allows you to get a call resource from a voice channel - all metadata is collected for further processing.
You can use Answer call in response to an incoming call, this will connect the call and allow you to respond or run additional steps in the flow while the call is ongoing.
Additionally, there are Advanced Configurations available, these options can be configured to customise the answer call rules to your liking:
Call ID: Identifies the call that will be connected to the destinations.
Voice Channel ID: The ChannelID that will be used to place the call. Defines the CallerID of the call.
Workspace Id: The Workspace ID of the call.
Wait For Conditions: Enable if you want to wait if the call is answered to proceed with the flow.
Terminate the channel call via Hangup call if the destination endpoint is either busy or rejects the call.
Additionally, there are Advanced Configurations available, these options can be configured to customise the hangup call rules to your liking:
Call ID: Identifies the call that will be connected to the destinations.
Voice Channel ID: The ChannelID that will be used to place the call. Defines the CallerID of the call.
Workspace Id: The Workspace ID of the call.
Wait For Conditions: Enable if you want to wait if the call is terminated to proceed with the flow.
You are able to add media files from URLs into the voice flow through the Play URL step.
Additionally, there are Advanced Configurations available, these options can be configured to customise the play URL rules to your liking:
Loop: Number of times allowed to play the defined media file.
Timeout: Maximum duration of the playback in seconds.
Call ID: Identifies the call that will be connected to the destinations.
Voice Channel ID: The ChannelID that will be used to place the call. Defines the CallerID of the call.
Workspace Id: The Workspace ID of the call.
Wait For Conditions: Enable if you want to wait until the announcement was successfully played to proceed with the flow.
Play a message during the call using text-to-speech, all you need to do is provide the text you want to play and set the language and desired voice experience (male, female).
Additionally, there are Advanced Configurations available, these options can be configured to customise the Say TTS rules to your liking:
Loop:Number of times allowed to play the announcement.
Timeout: Maximum duration of the say text-to-speech in seconds.
Call ID: Identifies the call that will be connected to the destinations.
Voice Channel ID: The ChannelID that will be used to place the call. Defines the CallerID of the call.
Workspace Id: The Workspace ID of the call.
Wait For Conditions: Enable if you want to wait if the announcement was successfully played to proceed with the flow.
The Gather Digits from a Call step facilitates input collection during calls (1-9, *, #). This is specially helpful when building an IVR and want the callee to select an option, press 1 to talk to sales, 2 to talk to support and so on. You have the option to play sounds or announcements defining the number of loops and timeouts or use the say text to call by selecting language (en-US), number of loops, text to be said, timeouts and voice experience (male, female).
Additionally, there are Advanced Configurations available, these options can be configured to customise the gather digits from a call rules to your liking:
End Key: Choose the input that concludes the digit gathering process. For instance, if you designate the End Key as # and the caller enters 1234#, the flow will cease waiting for further input once they press #.
Input: DTMF as default option.
Retries: Number of retries that are possible to insert input digits in case of failure.
Timeout: Maximum duration of the digits gathering process. Digits gathering is considered unsuccessful after the defined timeout in seconds.
Call ID: Identifies the call that will be connected to the destinations
Voice Channel ID: The ChannelID that will be used to place the call. Defines the CallerID of the call.
Workspace Id: The Workspace ID of the call.
Wait For Conditions: Enable if you want to wait if the digits were successfully collected to proceed with the flow.
The Record Call step allows you to start call recording from the caller's side. Recording can persist for a predefined duration or until a DTMF digit is pressed, offering end key options ranging from 1 to 9, as well as *, or #.
Additionally, there are Advanced Configurations available, these options can be configured to customise the record call rules to your liking:
End Key: Select the input to end the recording from 1 to 9, as well as *, or #.
Max Length: Maximum duration of the call recording in seconds.
Timeout: This attribute dictates the conclusion of recording after a specified duration of silence. To deactivate this function, set the timeout to 0.
Transcribe Locale: Choose the language for subsequent transcription (e.g., en-US).
Call ID: Identifies the call that will be connected to the destinations.
Voice Channel ID: The ChannelID that will be used to place the call. Defines the CallerID of the call.
Workspace Id: The Workspace ID of the call.
Wait For Conditions: Enable if you want to wait until the record is finished and is available to proceed with the flow.
Generate a transcription for a recording by choosing the previously collected recording ID variable and defining the desired language for the transcription (e.g. en-US).
When building a flow, it is possible to include the Get Recording step, allowing to gather the recording metadata or the corresponding file. When issuing a GET request to /recordings/:recordingId, you will receive the metadata of the recording. However, if you issue a GET request to /recordings/:recordingId.mp3, you will receive the actual audio file.
When building a flow, you can retrieve a call transcription using the Get Transcription step, you can store the transcription file in formats like .json, .vtt, .txt or .srt.