Before you start!
Confirm you have an Expert or Custom plan
Confirm CloudTalk Support has enabled speech-to-text for your account.
Also, don't forget to check out our main article on Workflow Automations!
Introduction
Speech-to-text transcription exporting can work differently for various types of calls. One reason that this can be particularly tricky is that both the triggers that are used to export a STT transcription do not always contain all of the required inputs.
This article will specifically discuss exporting speech-to-text for voicemail calls, if you're looking for general information on exporting speech-to-text files for standard calls, check out our general exporting guide here.
In order to identify voicemail calls we first need to understand the voicemail indicators used by the system. In simple terms, voicemail is essentially a missed call connected with a recording of the callers message. Whether the scenario triggering call is of the type 'missed' can be recognised by filtering calls with “talking_time equal to 0” (or any other reasonable threshold) and “direction equal to incoming”. With this in mind, we can start building our workflow scenario.
Getting started- locating your workflows
You will start by logging into your CloudTalk Dashboard. On the left side of the screen you will see a list of other tabs under Dashboard. To reach the automations tab, click on Account > Workflow Automations.
Click the blue button to + Add Workflow.
Note: you can give your new Workflow Automation a name to distinguish it from other automated tasks you set up.
Setting up your new speech-to-text voicemail workflow
A usual use case of retrieving a speech-to text-transcription from CloudTalk consists of 2 workflows:
Speech-to-text creation workflow
Transcription export workflow
1.) Create a regular speech to text creation scenario with the following:
Trigger: Recording uploaded
Action: Speech to text
2.) Add a condition before your speech to text action with the following parameters in order to isolate voicemail calls:
event.properties.talking_time is equal to 0
event.properties.direction is equal to incoming
Note- You can add further conditions to the same condition node to filter out certain call recordings from transcription if you wish.
3.) Save and activate your scenario
4.) Create a second workflow for exporting with the following:
Trigger: Transcription created
Action: According to your needs (how and where you'd like to export your data e.g sms/slack)
Exporting to Slack
If you're looking for information on exporting to Slack specifically, check out the second half of our general speech-to-text exporting article.
Exporting via SMS
In case you'd like to use the SMS Action type to export speech-to-text via this method, the setup additionally consists of 5 required parameters:
Sender - A phone number, usually represented by the agent’s number. Outbound SMSs must be supported and the number must be in E164 format.
Recipient - Phone number of the contact which the SMS should be delivered to. The number must be in E164 format.
SMS text - Any text that you'd like to be displayed within the body of the text message before you receive your transcription.
Authorization - ACCESS KEY ID and ACCESS KEY SECRET
5.) Save and activate your scenario.
If you have any further questions please do not hesitate to reach out to our Support team, we are always happy to help!