Get questions

The get questions endpoint identifies specific tasks resulting from the conversation that require action. Usually, these action items require one or more people to do something, such as schedule a meeting or complete a task. This endpoint is useful for task tracking and ensuring follow-through on commitments made during a conversation.

Our AI model:

• Identifies directed questions, e.g., "Does John have access to this repository?"
• Detects implicit and actionable questions, e.g., "do you think" or "Should Ana make this call?"
• Excludes casual small talk and irrelevant questions.

An example could be, "When is the project deadline?" or "Who is responsible for the marketing campaign?".

Before we start

Transcripts must be generated upon request. If you have already generated a transcript, you can use the recordingId to get the questions.

How to use

You can use the following cURL to obtain a response on the endpoint:

curl --location --request GET https://api.superviz.com/recordings/transcripts/{recordingId}/questions \
  --header 'Content-Type: application/json' \
  --header "client_id: ${YOUR CLIENT_ID}" \
  --header "secret: ${YOUR SECRET}" \
  --header "passphrase: ${YOUR PASSPHRASE}"

HTTP Request

GET https://api.superviz.com/recordings/transcripts/{recordingId}/questions

Name	Description
recordingId	This field contains the unique identification of the recording, which can be found with the GET Recordings endpoint.

Headers

When using this endpoint, you need to provide the Authorization Bearer token and the passphrase. The following headers are required:

Name	Description
client_id	Required. You need to use the Client ID to authenticate in this repository. You can retrieve your Client ID under the Dashboard > Developer > Keys.
secret	Required. You need to use the Secret Key to authenticate in this repository. You can create a new API Secret under our Dashboard > Developer > Keys.
passphrase	Required. The passphrase of your organization that allows the decryption of the transcript content. You can create a new passphrase under our Dashboard > Developer > Keys.

Response

Status code 200 will indicate that the request was successful. The response will be a list of objects in JSON format and include the following fields:

Name	Type	Description
text	string	This field contains one of the questions that were asked during the meeting.
userName	string	This field contains the number of the current speaker. Values will be always Speaker # where the # symbolizes which speaker it is. NOTE: It does not represent the participant.name provided when initializing a meeting. Example: In a meeting in which three participants spoke, the first one who spoke would be Speaker 1, Speaker 2 for the second voice, and Speaker 3 would be the one who spoke last.
score	number	This field contains the score of the question.

Example:

[
	{
		"text": "...",
		"userName": "...",
		"score": 0.97
	},
	...
]