Provisioning Configuration
Provisioning chains orchestrate voice interaction pipelines by linking Speech-to-Text (STT), Agent Configuration, and Text-to-Speech (TTS) models. Used for voice-based deployments with CHAINED provisioning type.
Source: src/core/service-configuration/provisioning-config.ts
SttModelConfig​
Speech-to-Text model configuration for the input stage.
Schema: SttModelConfigSchema
Attributes​
| Attribute | Type | Required | Default | Description |
|---|---|---|---|---|
modelId | string | Yes | — | STT model ID from WIIL registry |
defaultLanguage | string | No | en-US | Default language for speech recognition |
TtsModelConfig​
Text-to-Speech model configuration for the output stage.
Schema: TtsModelConfigSchema
Attributes​
| Attribute | Type | Required | Default | Description |
|---|---|---|---|---|
modelId | string | Yes | — | TTS model ID from WIIL registry |
voiceId | string | Yes | — | Voice ID for speech synthesis |
defaultLanguage | string | No | en-US | Default language for synthesis |
voiceSettings | object | No | — | Voice-specific settings (pitch, speed, etc.) |
ProvisioningConfigChain​
Complete voice interaction processing pipeline.
Schema: ProvisioningConfigChainSchema
Pipeline Flow​
User Speech → STT → Text → Agent → Text Response → TTS → Agent Speech
Attributes​
| Attribute | Type | Required | Default | Description |
|---|---|---|---|---|
id | string | — | — | Unique identifier for the chain |
chainName | string | Yes | — | Human-readable name |
description | string | No | — | Chain purpose and configuration details |
sttConfig | SttModelConfig | Yes | — | Speech-to-text configuration (stage 1) |
agentConfigurationId | string | Yes | — | Agent configuration ID (stage 2) |
ttsConfig | TtsModelConfig | Yes | — | Text-to-speech configuration (stage 3) |
createdAt | number | — | — | Unix timestamp when created |
updatedAt | number | — | — | Unix timestamp when last updated |
TranslationChainConfig​
Extended chain for real-time language translation.
Schema: TranslationChainConfigSchema
Pipeline Flow​
Speech (Lang A) → STT → Text (Lang A) → Translation → Text (Lang B) → TTS → Speech (Lang B)
Additional Attributes​
| Attribute | Type | Required | Default | Description |
|---|---|---|---|---|
processingModelId | string | Yes | — | LLM model ID for translation |
isTranslation | boolean | No | true | Flag indicating translation chain |
Example​
Provisioning Chain​
{
"id": "chain_abc123",
"chainName": "Customer Support Voice Chain",
"description": "Voice processing chain for customer support calls",
"sttConfig": {
"modelId": "whisper-v3",
"defaultLanguage": "en-US"
},
"agentConfigurationId": "agent_789",
"ttsConfig": {
"modelId": "eleven-labs-v2",
"voiceId": "adam",
"defaultLanguage": "en-US",
"voiceSettings": {
"stability": 0.75,
"similarity_boost": 0.5
}
},
"createdAt": 1709856000000,
"updatedAt": 1709856000000
}
Translation Chain​
{
"id": "chain_def456",
"chainName": "EN-ES Translation Chain",
"description": "English to Spanish translation for customer support",
"sttConfig": {
"modelId": "whisper-v3",
"defaultLanguage": "en-US"
},
"processingModelId": "gpt-4-translator",
"ttsConfig": {
"modelId": "eleven-labs-v2",
"voiceId": "spanish-voice",
"defaultLanguage": "es-ES"
},
"isTranslation": true,
"createdAt": 1709856000000,
"updatedAt": 1709856000000
}
Create / Update Schemas​
| Schema | Description | Omits |
|---|---|---|
CreateProvisioningConfigSchema | For creating chains | id, createdAt, updatedAt |
UpdateProvisioningConfigSchema | For partial updates | All fields optional except id |
CreateTranslationChainConfigSchema | For creating translation chains | id, createdAt, updatedAt |
UpdateTranslationChainConfigSchema | For updating translation chains | All fields optional except id |
Related​
- Deployment Configuration - References chains via
provisioningConfigChainId - Agent Configuration - Referenced by chains for processing
- Voice & Language - Voice and language options