Set up chained configuration for Voice

In this guide, you'll learn how to set up chained configuration. For information on ingestion modes and supported channels, see Cross-channel support.

What is chained configuration?

(information)

Info

Chained configuration is available for post-call batch transcription only. For real-time transcription, see Real-time transcription and TwiML.

Chained configuration automates the workflow from audio capture to AI-driven insights. By linking specific API resources, you can eliminate manual handoffs and automatically trigger Conversation Orchestrator and Conversation Intelligence for post-call batch transcription.

Post-call batch transcription

(information)

Batch transcription workflow not supported in Conversation Intelligence (classic)

The Transcription Configuration resource for batch transcription isn't available for customers using Conversation Intelligence (classic). To use it with the latest version of Conversation Intelligence, contact Support or your Twilio account team.

Post-call batch transcription lets you process recordings through Conversation Orchestrator and Conversation Intelligence after a conversation ends. Chain a recording configuration, transcription configuration, and conversation configuration to automatically trigger analysis on completed recordings.

To set up a chained configuration for post-call batch transcription, follow these steps:

Create required configurations in the following order:
Twilio ConsoleAPI
Create a conversation configuration without capture rules.

In the Twilio Console, go to Orchestration > Conversation Orchestrator > Conversation configurations.

Click Create a conversation configuration and follow the steps in the UI.

Skip Messaging and chat traffic.

Skip Voice traffic to ensure no capture rules are created.

Create an intelligence configuration and link it to the conversation configuration.

In the Twilio Console, go to Orchestration > Conversation Intelligence > Intelligence configurations.

Click Create Intelligence configuration and follow the steps in the UI.
Make sure to attach the conversation configuration you created in the previous step.

Click Create rule for the intelligence configuration and follow the steps in the UI.

Add language operators and set parameters.

Apply triggers and actions. Note: For batch transcription, you must use the At conversation end trigger.

Optionally, enable context.

For more information, see Define rules.

Create a transcription configuration.

In the Twilio Console, go to Communications > Voice > Transcriptions.

Click Create transcription and follow the steps in the UI.

To receive the transcription results to your application using a webhook, set Transcription status callback.

Optionally, you can pass your conversation configuration ID.

If you pass a configuration ID, it automatically activates Conversation Orchestrator and Conversation Intelligence when the transcript is ready.

If you don't pass a configuration ID, the transcript is ephemeral and only accessible through the webhook.

Create a recording configuration.

In the Twilio Console, go to Communications > Voice > Recordings > Configurations.

Click Create recording configuration and follow the steps in the UI.

To receive recording status events, set Call recording status callback.

Make sure to attach the transcription configuration you created in the previous step.
Create a conversation configuration without capture rules.

Create a conversation configuration for post-call batch transcription
1
// Download the helper library from https://www.twilio.com/docs/node/install
2
const twilio = require("twilio"); // Or, for ESM: import twilio from "twilio";
3

4
// Find your Account SID at twilio.com/console
5
// Provision API Keys at twilio.com/console/runtime/api-keys
6
// and set the environment variables. See http://twil.io/secure
7
// For local testing, you can use your Account SID and Auth token
8
const accountSid = process.env.TWILIO_ACCOUNT_SID;
9
const apiKey = process.env.TWILIO_API_KEY;
10
const apiSecret = process.env.TWILIO_API_SECRET;
11
const client = twilio(apiKey, apiSecret, { accountSid: accountSid });
12

13
async function createConfiguration() {
14
const configuration = await client.conversations.v2.configurations.create({
15
displayName: "post-call-batch-configuration",
16
description: "Intelligence configuration for post-call batch transcription",
17
conversationGroupingType: "GROUP_BY_PROFILE",
18
memoryStoreId: "<MEMORY_STORE_ID>",
19
channelSettings: {
20
VOICE: {
21
statusTimeouts: {
22
inactive: 30,
23
closed: 1440,
24
},
25
},
26
},
27
});
28

29
console.log(configuration.statusUrl);
30
}
31

32
createConfiguration();

Create an intelligence configuration and link it to the conversation configuration.

Make a POST request to the Intelligence Configurations resource.

Note: For batch transcription, you must use the CONVERSATION_END trigger.

The following example uses the Twilio-authored Summary operator to send a summary to a webhook when the conversation ends.

Create an intelligence configuration
1
// Download the helper library from https://www.twilio.com/docs/node/install
2
const twilio = require("twilio"); // Or, for ESM: import twilio from "twilio";
3

4
// Find your Account SID at twilio.com/console
5
// Provision API Keys at twilio.com/console/runtime/api-keys
6
// and set the environment variables. See http://twil.io/secure
7
// For local testing, you can use your Account SID and Auth token
8
const accountSid = process.env.TWILIO_ACCOUNT_SID;
9
const apiKey = process.env.TWILIO_API_KEY;
10
const apiSecret = process.env.TWILIO_API_SECRET;
11
const client = twilio(apiKey, apiSecret, { accountSid: accountSid });
12

13
async function createConfiguration() {
14
const configuration = await client.intelligence.v3.configurations.create({
15
displayName: "post-call-batch-intelligence",
16
description: "Intelligence configuration for post-call batch transcription",
17
rules: [
18
{
19
operators: [
20
{
21
id: "intelligence_operator_01kcv35pnkeysaf6z6cqtbpegn",
22
},
23
],
24
triggers: [
25
{
26
on: "CONVERSATION_END",
27
},
28
],
29
actions: [
30
{
31
type: "WEBHOOK",
32
method: "POST",
33
url: "https://crm.example.com/conversation_summary",
34
},
35
],
36
context: {
37
memory: {
38
enabled: true,
39
},
40
},
41
},
42
],
43
});
44

45
console.log(configuration.accountId);
46
}
47

48
createConfiguration();

To attach the intelligence configuration, make a PUT request to the Conversations Configuration resource.

Attach your intelligence configuration to a conversation configuration
1
// Download the helper library from https://www.twilio.com/docs/node/install
2
const twilio = require("twilio"); // Or, for ESM: import twilio from "twilio";
3

4
// Find your Account SID at twilio.com/console
5
// Provision API Keys at twilio.com/console/runtime/api-keys
6
// and set the environment variables. See http://twil.io/secure
7
// For local testing, you can use your Account SID and Auth token
8
const accountSid = process.env.TWILIO_ACCOUNT_SID;
9
const apiKey = process.env.TWILIO_API_KEY;
10
const apiSecret = process.env.TWILIO_API_SECRET;
11
const client = twilio(apiKey, apiSecret, { accountSid: accountSid });
12

13
async function updateConfiguration() {
14
const configuration = await client.conversations.v2
15
.configurations("<CONVERSATION_CONFIGURATION_ID>")
16
.update({
17
displayName: "post-call-batch-configuration",
18
description:
19
"Conversation configuration for post-call batch transcription",
20
conversationGroupingType: "GROUP_BY_PROFILE",
21
memoryStoreId: "<MEMORY_STORE_ID>",
22
channelSettings: {
23
VOICE: {
24
statusTimeouts: {
25
inactive: 30,
26
closed: 1440,
27
},
28
},
29
},
30
intelligenceConfigurationIds: ["<INTELLIGENCE_CONFIGURATION_ID>"],
31
});
32

33
console.log(configuration.statusUrl);
34
}
35

36
updateConfiguration();

Create a transcription configuration.

To receive the transcription results to your application using a webhook, set transcriptionStatusCallback.

Optionally, you can pass a conversationConfigurationId.

If you pass a conversationConfigurationId, it automatically activates Conversation Orchestrator and Conversation Intelligence when the transcript is ready.

If you don't pass a conversationConfigurationId, the transcript is ephemeral and only accessible via the transcript finalization webhook.

The following example uses the Twilio-managed automatic speech model. For other engine and language combinations, see Create a transcription configuration.

1
curl -X POST "https://voice.twilio.com/v2/Configurations/Transcription" \
2
-u "$TWILIO_ACCOUNT_SID:$TWILIO_AUTH_TOKEN" \
3
-H "Content-Type: application/json" \
4
-d '{
5
"unique_name": "transcription_with_twilio_managed",
6
"description": "Transcription configuration using Twilio Managed engine",
7
"configuration": {
8
"configurationType": "Transcription",
9
"transcriptionEngine": "twilio_managed",
10
"language": "en-US",
11
"transcriptionStatusCallback": {
12
"url": "https://example.com/transcription-callback",
13
"method": "POST"
14
},
15
"participantDefaults": [
16
{ "audioChannelIndex": 1, "type": "CUSTOMER" },
17
{ "audioChannelIndex": 2, "type": "HUMAN_AGENT" }
18
],
19
"conversationConfigurationId": "<CONVERSATION_CONFIGURATION_ID>"
20
}
21
}'

Create a recording configuration.

To receive recording status events, set callRecordingStatusCallback.

To link the recording configuration to the transcription configuration, set features.type to transcriptionConfiguration and features.featureId to the transcription configuration ID.

1
curl -X POST "https://voice.twilio.com/v2/Configurations/Recording" \
2
-u "$TWILIO_ACCOUNT_SID:$TWILIO_AUTH_TOKEN" \
3
-H "Content-Type: application/json" \
4
-d '{
5
"unique_name": "my_recording_config",
6
"description": "My recording configuration",
7
"configuration": {
8
"configurationType": "Recording",
9
"compositionPolicy": {
10
"channels": "dual",
11
"trim": "trim-silence"
12
},
13
"callRecordingStatusCallback": {
14
"url": "https://example.com/call-recording-status",
15
"method": "POST",
16
"events": ["in-progress", "completed"]
17
},
18
"features": [
19
{
20
"type": "transcriptionConfiguration",
21
"featureId": "<TRANSCRIPTION_CONFIGURATION_ID>"
22
}
23
]
24
}
25
}'
Submit recordings for transcription.
New callsExisting recordings
For new calls, link the recording configuration so Twilio automatically transcribes every recording it produces.
To use a recording configuration, pass the configuration ID as the recordingConfigurationId parameter in any of the following recording methods:

<Record> in TwiML

<Dial record> in TwiML

<Conference record> in TwiML

<Record=true> in an outbound call using the REST API

POST to Recording resource of an in-progress Call SID

<Start><Recording> in TwiML
For existing recordings, submit them directly so Twilio hydrates Conversation Orchestrator and fires your Conversation Intelligence rules once the transcript is ready.
To submit an existing recording for transcription, make a POST request to the Transcriptions resource with the following:

transcriptionConfigurationId: The ID of the transcription configuration you created, with a conversationConfigurationId attached.

sourceId: The Recording SID (RE...) of the existing recording.

Transcribe an existing recording
1
// Download the helper library from https://www.twilio.com/docs/node/install
2
const twilio = require("twilio"); // Or, for ESM: import twilio from "twilio";
3

4
// Find your Account SID at twilio.com/console
5
// Provision API Keys at twilio.com/console/runtime/api-keys
6
// and set the environment variables. See http://twil.io/secure
7
// For local testing, you can use your Account SID and Auth token
8
const accountSid = process.env.TWILIO_ACCOUNT_SID;
9
const apiKey = process.env.TWILIO_API_KEY;
10
const apiSecret = process.env.TWILIO_API_SECRET;
11
const client = twilio(apiKey, apiSecret, { accountSid: accountSid });
12

13
async function createV3Transcriptions() {
14
const transcription = await client.voice.v3.transcriptions.create({
15
transcriptionConfigurationId:
16
"voice_transcriptionconfiguration_5pe8jw3ahdmsh7zr06yh4d45x1",
17
sourceId: "YOUR_RECORDING_ID",
18
});
19

20
console.log(transcription.status);
21
}
22

23
createV3Transcriptions();
Test the configuration.
New callsExisting recordings
Make a call to your Twilio number and complete the conversation.

After the call ends, you'll have the following:

A recording of the call

A transcription of the call

A conversation in Conversation Orchestrator

A result of the configured intelligence rules, such as a summary of the conversation
Submit the existing recording for transcription using the steps above.

Once the transcription's status is COMPLETED, you'll have the following:

A transcription of the recording

A conversation in Conversation Orchestrator

A result of the configured intelligence rules, such as a summary of the conversation

Set up chained configuration for Voice

What is chained configuration?

Info

Post-call batch transcription

Batch transcription workflow not supported in Conversation Intelligence (classic)

Resources