Workflow Automations are designed to make the job of an agent easier by automating processes such as sending a followup text, transcribing certain calls, or automatically forwarding transcriptions to Slack. The endpoint of the automation process is called an Action (not to be confused with an Object Action assigned to a Condition or Trigger). In a Workflow, Actions come as the result of a Trigger meeting our specified Conditions.
Setting Up an Action
In the Workflow Automations tab under Settings, we are given the option to + Add Workflow. After we have named our new Workflow, selected a Trigger, and possibly added some Conditions, we are given the option to Add Action.
Note: You can add more than one Action for the same Trigger. In cases
where differing Triggers are required, it is better to use a combination
of more than one Workflow to create the desired scenario.
CloudTalk currently supports four Action types (open in new tab):
Speech to Text Transcription (if the Trigger is Recording Uploaded)
Sending an API Request
This Action type is good for tech-savvy users looking for more creative freedom in designing automations. An API (Application Programming Interface) allows different web applications to communicate. At this time, requests can be made either to the CloudTalk REST API or other, external systems which use:
Authentication schemes requiring static, immutable headers (e.g., Basic or Bearer authentication schemes)
Systems with URL / JSON body authentication.
CloudTalk's API is one example of the latter type. Each API Action consists of the following components:
"Name" | Definition | Example Value |
Endpoint | URL of the respective API endpoint |
|
Method | Defines the action of the API request | Accepted: |
Headers | Metadata, often necessary for authentication and other purposes | Key: |
Values | Define the body of the request in the form of a JSON object | Key: Value: Format: |
Use the Values section of the API request builder to define the body of a request. Each entry in the Values section will contain a Key, Value, and Format.
In this instance, Key is the property name in the API request. For example:
contactName
Value can be static. It might be something like:
Johnny Snow
This value could also be a dynamic property which sends the value input based on the data taken from your chosen Trigger. To achieve a dynamic value, choose from one of the blue phrases in the Useful data for your Workflow sidebar. As shown below, you should see the Value box populate with something like:
{{event.properties.name}}
This example would pull the name value from the contact associated with your Trigger.
Format value describes a data type, such that the target system (API) can correctly interpret any enclosed data. CloudTalk supports the following formats:
Number - whole (integers) or decimals (floats with a dot decimal separator)
String - sequence of alphanumeric characters
Boolean - evaluates to True or False (any undefined form defaults as False)
Auto - type is detected automatically
Sending an SMS
This Action type can be used to send a followup text to a client, or to notify internal team members of important activity. There are five required parameters for sending an SMS:
Sender
> CloudTalk number in E164 format from which the SMS should be sent (outbound SMS must be supported)
Recipient
> Phone number of the contact who should receive the SMS, in E164 format
SMS text
> Text for the body of the message
Authorization
> ACCESS KEY ID and ACCESS KEY SECRET (must be obtained by an admin)
As with API requests, the value of each parameter can be represented by a static entry or by a dynamic value, obtained by clicking into the Value box and then clicking on one of the blue phrases in the Useful data for your Workflow sidebar.
Sending a Notification to Slack
This Action type is more popular for internal notifications, where important information can be forwarded to a company Slack channel. When choosing this option, make sure you have a connection to a respective Slack workspace and channel. Also ensure that Token scopes for this workspace and channel include chat:write permissions. Parameters are:
Token
> an authentication Token generated by Slack's interface
Channel
> Slack channel where messages will be posted
Text
> message body
Icon_emoji
Icon_url
Use_markdown
> when True, **text** is shown as text
As with API requests, the Value of each parameter can be represented by a static entry or by a dynamic value, obtained by clicking into the Value box and then clicking on one of the blue phrases in the Useful data for your Workflow sidebar.
The resulting Slack notification might look something like this:
Speech to Text Transcription
This Action type is often used to store the written form of conversations recorded during a call. Be aware that this Action is only enabled after choosing Recording Uploaded as a Trigger. There are three required Keys for Speech to Text Transcription:
> Note that some languages are supported by additional models which are optimized for certain audio types: phone call, enhanced phone call, and enhanced video. If any of these additional models are available for your language, they will be listed within the default and command and search models for your language. If only the regular default and command and search models are listed, that means no additional models are currently available for your language.
> Models have been optimized for accuracy (for an increased cost)
> Choose a specific machine learning model (
phone_callis usually sufficient)
If you have any further questions, please reach out to our Support team, we are always happy to help!