Create assistant

Creates a new AI assistant. The assistant is created with inactive status and is ready to configure and activate.

Endpoint

POST /user/assistant

Engine modes

The assistant supports three engine modes, each with different requirements:

Mode	Description	Required field
`pipeline`	Traditional STT → LLM → TTS pipeline	`llm_model_id`
`multimodal`	Real-time multimodal AI	`multimodal_model_id`
`dualplex`	Multimodal brain with custom TTS voice	`multimodal_model_id`

Request

Core required fields

name

string

required

Name of the assistant. Maximum 255 characters.

voice_id

integer

required

Voice ID to use. Use the Get Voices endpoint with the mode parameter to retrieve voices compatible with your engine mode.

language_id

integer

required

Primary language ID. Use the Get Languages endpoint to retrieve available languages.

type

string

required

Assistant type: inbound or outbound.

mode

string

required

Engine mode: pipeline, multimodal, or dualplex.

timezone

string

required

Timezone identifier (e.g., America/New_York, Europe/Bucharest).

initial_message

string

required

Greeting message spoken at the start of a call. Maximum 200 characters.

system_prompt

string

required

System prompt defining the assistant’s behavior and personality.

Mode-specific fields

llm_model_id

integer

LLM model ID. Required for pipeline mode. Use the Get Models endpoint to retrieve available models.

multimodal_model_id

integer

Multimodal model ID. Required for multimodal and dualplex modes. Use the Get Models endpoint to retrieve available multimodal models.

chat_llm_fallback_id

integer

Fallback LLM model ID for tool calls in multimodal/dualplex modes. Optional.

turn_detection_threshold

number

Turn detection sensitivity for multimodal/dualplex modes. Range: 0–1. Defaults to auto.

Secondary languages

secondary_language_ids

integer[]

Additional language IDs the assistant can speak. The assistant auto-detects and switches languages during the call.

"secondary_language_ids": [2, 3, 4]

Knowledge base

knowledgebase_id

integer

ID of the knowledge base to attach to this assistant.

knowledgebase_mode

string

How the assistant uses the knowledge base:

function_call — AI calls a function to search. Required for multimodal and dualplex modes.
prompt — Knowledge is injected into the system prompt. Pipeline mode only.

Phone number

phone_number_id

integer

ID of the phone number to assign. Must belong to your account.

For inbound assistants, the phone number cannot be a Caller ID type and cannot already be assigned to another inbound assistant.

Custom mid-call tools

tool_ids

integer[]

IDs of custom mid-call tools to attach. Each tool must belong to your account.

"tool_ids": [1, 5, 12]

Built-in tools

tools

array

Built-in tools to enable on the assistant. Each tool object has a type and type-specific fields.

Show Available tool types

call_transfer — Transfer the call to another number.

phone_number (required): Destination phone number (e.g., +1234567890).
description: When the AI should transfer.
custom: If true, the AI dynamically determines the transfer number.
warm_transfer: Send a message before transferring. Default: false.
warm_transfer_message: Prompt for what the AI says before transferring.

warm_call_transfer — Warm transfer with supervisor briefing.

supervisor_phone (required): Phone number to dial for the transfer.
outbound_phone_id (required): ID of the phone number used to dial the supervisor.
description (required): When the AI should initiate the warm transfer.
caller_id_mode: Number the supervisor sees: outbound_number (default), customer_number, or custom.
hold_music: Hold audio: hold_music (default) or none.
hold_message: Message spoken to the caller before being placed on hold.

end_call — End the call programmatically.

description: When the AI should end the call.

dtmf_input — Send DTMF tones for IVR navigation.

description: When to use DTMF input.

collect_keypad — Collect keypad input from the caller.

timeout: Seconds to wait for input. Range: 1–30. Default: 5.
stop_key: Key that ends input: # (default) or *.

calendar_integration — Schedule appointments via Cal.com.

calcom_api_key (required): Your Cal.com API key.
calcom_event_slug (required): Event type slug from Cal.com.
description: When the AI should offer scheduling.

Voice and TTS settings

voice_stability

number

default:"0.70"

Voice stability. Range: 0–1. Higher values produce a more consistent voice.

voice_similarity

number

default:"0.50"

Voice similarity to the original. Range: 0–1.

speech_speed

number

default:"1.00"

Speech speed multiplier. Range: 0.7–1.2.

llm_temperature

number

default:"0.10"

LLM temperature. Range: 0–1. Lower values produce more deterministic responses.

tts_emotion_enabled

boolean

default:"true"

Whether to enable emotional TTS synthesis.

synthesizer_provider_id

integer

Custom TTS provider ID. Auto-selected based on language if not provided.

transcriber_provider_id

integer

Custom STT provider ID. Pipeline mode only. Auto-selected based on language if not provided.

Call behavior

allow_interruptions

boolean

default:"true"

Whether the caller can interrupt the assistant.

Cannot be disabled for multimodal and dualplex modes.

fillers

boolean

default:"false"

Whether to use filler audio while processing (e.g., “Hmm”, “Let me check”).

Only available in pipeline mode.

filler_config

object

Custom filler phrases per category. Each category is an array of short strings.

"filler_config": {
  "positive": ["Great!", "Perfect!"],
  "negative": ["Hmm.", "I see."],
  "question": ["Good question.", "Let me check."],
  "neutral": ["Ok.", "Noted."]
}

record

boolean

default:"false"

Whether to record calls.

enable_noise_cancellation

boolean

default:"true"

Whether to enable noise cancellation.

wait_for_customer

boolean

default:"false"

If true, the assistant waits for the customer to speak first.

Timing

max_duration

integer

default:"600"

Maximum call duration in seconds. Range: 20–1200.

max_silence_duration

integer

default:"40"

Maximum silence duration before re-engagement, in seconds. Range: 1–360.

max_initial_silence_duration

integer

Maximum silence at call start before ending, in seconds. Range: 1–120.

ringing_time

integer

default:"30"

Maximum ringing time before giving up, in seconds. Range: 1–60.

Re-engagement

reengagement_interval

integer

default:"30"

Seconds between re-engagement attempts. Range: 7–600.

reengagement_prompt

string

Custom re-engagement message prompt. Maximum 1000 characters.

Voicemail

end_call_on_voicemail

boolean

default:"true"

Whether to end the call when voicemail is detected.

voice_mail_message

string

Message to leave on voicemail before hanging up. Maximum 1000 characters.

Endpoint detection

endpoint_type

string

default:"vad"

Voice activity detection type: vad or ai.

endpoint_sensitivity

number

default:"0.5"

Endpoint sensitivity level. Range: 0–5.

interrupt_sensitivity

number

default:"0.5"

Interrupt sensitivity level. Range: 0–5.

min_interrupt_words

integer

Minimum words spoken before an interruption is processed. Range: 0–10.

Ambient sound

ambient_sound

string

Background ambient sound: off, office, city, forest, crowded_room, cafe, or nature.

ambient_sound_volume

number

default:"0.5"

Ambient sound volume. Range: 0–1.

Webhook configuration

is_webhook_active

boolean

default:"false"

Whether to enable post-call webhook notifications.

webhook_url

string

Webhook URL for post-call notifications. Required if is_webhook_active is true.

send_webhook_only_on_completed

boolean

default:"true"

Whether to send webhooks only for completed calls.

include_recording_in_webhook

boolean

default:"true"

Whether to include the recording URL in webhook payloads.

Post-call evaluation

post_call_evaluation

boolean

default:"true"

Whether to enable AI post-call evaluation.

post_call_schema

array

Schema for extracting structured data from calls.

Show Schema field properties

name

string

required

Field name. 3–16 characters, lowercase alphanumeric and underscores only.

type

string

required

Data type: string, number, or bool.

description

string

required

Description of the field. 3–255 characters.

"post_call_schema": [
  {"name": "status", "type": "bool", "description": "Was the call objective achieved"},
  {"name": "summary", "type": "string", "description": "Brief summary of the call"}
]

Variables

variables

object

Custom key-value pairs accessible in prompts via {{variable_name}}.

"variables": {
  "company_name": "Acme Corp",
  "product": "Premium Widget"
}

Conversation ended settings

conversation_inactivity_timeout

integer

default:"30"

Minutes of chat inactivity before the conversation is considered ended. Range: 1–1440.

conversation_ended_retrigger

boolean

default:"false"

Whether to allow re-triggering after the conversation ends due to inactivity.

conversation_ended_webhook_url

string

Webhook URL called when a chat conversation ends due to inactivity.

Response

message

string

Confirmation message.

data

object

Show properties

integer

Unique identifier of the created assistant.

name

string

Name of the assistant.

status

string

Status — always inactive for new assistants.

type

string

inbound or outbound.

mode

string

Engine mode: pipeline, multimodal, or dualplex.

Example

curl -X POST "https://portal.intellixent.ai/api/user/assistant" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Sales Assistant",
    "voice_id": 1,
    "language_id": 1,
    "type": "outbound",
    "mode": "pipeline",
    "timezone": "America/New_York",
    "initial_message": "Hello! How can I help you today?",
    "system_prompt": "You are a professional sales assistant...",
    "llm_model_id": 2,
    "record": true,
    "tools": [
      {
        "type": "end_call",
        "description": "End call when customer is satisfied"
      }
    ]
  }'

201 Success

{
  "message": "Assistant created successfully",
  "data": {
    "id": 789,
    "name": "Sales Assistant",
    "status": "inactive",
    "type": "outbound",
    "mode": "pipeline"
  }
}

422 Validation error

{
  "message": "Validation failed",
  "errors": {
    "name": ["The name field is required."],
    "voice_id": ["The selected voice is not compatible with the chosen engine type."]
  }
}

New assistants are created with inactive status. Activate them through the Intellixent portal or by attaching a phone number and enabling them.

Overview

Assistants

Calls

Campaigns & Leads

Conversations

Messaging

Phone Numbers

Knowledge & Tools

Webhooks

Endpoint

Engine modes

Request

Core required fields

Mode-specific fields

Secondary languages

Knowledge base

Phone number

Custom mid-call tools

Built-in tools

Voice and TTS settings

Call behavior

Timing

Re-engagement

Voicemail

Endpoint detection

Ambient sound

Webhook configuration

Post-call evaluation

Variables

Conversation ended settings

Response

Example

Overview

Assistants

Calls

Campaigns & Leads

Conversations

Messaging

Phone Numbers

Knowledge & Tools

Webhooks

​Endpoint

​Engine modes

​Request

​Core required fields

​Mode-specific fields

​Secondary languages

​Knowledge base

​Phone number

​Custom mid-call tools

​Built-in tools

​Voice and TTS settings

​Call behavior

​Timing

​Re-engagement

​Voicemail

​Endpoint detection

​Ambient sound

​Webhook configuration

​Post-call evaluation

​Variables

​Conversation ended settings

​Response

​Example

Endpoint

Engine modes

Request

Core required fields

Mode-specific fields

Secondary languages

Knowledge base

Phone number

Custom mid-call tools

Built-in tools

Voice and TTS settings

Call behavior

Timing

Re-engagement

Voicemail

Endpoint detection

Ambient sound

Webhook configuration

Post-call evaluation

Variables

Conversation ended settings

Response

Example