.do
ApiAutomation

VapiAssistant

A voice AI assistant configuration for Vapi, supporting conversational AI with voice synthesis, transcription, LLM integration, and tool execution.

VapiAssistant

A voice AI assistant configuration for Vapi, supporting conversational AI with voice synthesis, transcription, LLM integration, and tool execution.

Properties

  • id (string): Unique identifier for the assistant.
  • name (string): Name of the voice assistant.
  • business ([[Business]]): The business that owns this assistant.
  • description (string): Description of assistant purpose and capabilities.
  • status (string): Assistant status (active, inactive, training, archived).
  • firstMessage (string): Initial greeting message when call starts.
  • systemPrompt ([[VapiPrompt]]): System prompt defining assistant behavior and personality.
  • model ([[AssistantModel]]): LLM model configuration for the assistant.
  • voice ([[VapiVoice]]): Voice configuration for speech synthesis.
  • transcriber ([[VapiTranscriber]]): Transcription service configuration for speech-to-text.
  • tools ([[VapiTool]][]): Available tools and functions the assistant can execute.
  • knowledgeBase ([[VapiKnowledgeBase]]): Knowledge base for contextual information.
  • language (string): Primary language code (en, es, fr, de, etc.).
  • supportedLanguages (string[]): All supported language codes.
  • maxCallDuration (number): Maximum call duration in seconds.
  • silenceTimeout (number): Timeout for user silence in seconds before ending call.
  • interruptSensitivity (number): Sensitivity for interruptions (0-1, higher = more sensitive).
  • endCallOnSilence (boolean): Whether to end call after silence timeout.
  • endCallPhrases (string[]): Phrases that trigger call ending (goodbye, hang up, etc.).
  • forwardingNumber ([[VapiPhoneNumber]]): Phone number to forward calls to if needed.
  • voicemailDetection (boolean): Whether to detect voicemail and handle accordingly.
  • voicemailMessage (string): Message to leave if voicemail is detected.
  • recordCalls (boolean): Whether to record calls.
  • transcribeCalls (boolean): Whether to generate call transcripts.
  • analyticsSummary (boolean): Whether to generate analytics summaries.
  • hipaaCompliant (boolean): Whether assistant is HIPAA compliant.
  • privacyMode (boolean): Enhanced privacy mode for sensitive conversations.
  • backgroundSound (string): Background sound option (office, cafe, none).
  • backgroundSoundLevel (number): Background sound volume (0-1).
  • serverUrl ([[URL]]): Webhook URL for receiving call events.
  • serverSecret (string): Secret for validating webhook requests.
  • metadata (object): Custom key-value pairs for additional data.
  • totalCalls (number): Total number of calls handled by this assistant.
  • averageCallDuration (number): Average call duration in seconds.
  • successRate (number): Percentage of successful calls (0-100).
  • created (number): Timestamp when assistant was created (Unix timestamp).
  • updated (number): Timestamp of last update (Unix timestamp).

Actions

create

Create a new voice assistant.

Input: [[VapiAssistant]]
Output: [[VapiAssistant]]

update

Modify assistant configuration.

Input: [[VapiAssistant]]
Output: [[VapiAssistant]]

delete

Delete the assistant.

Input: [[VapiAssistant]]
Output: [[VapiAssistant]]

activate

Activate the assistant for handling calls.

Input: [[VapiAssistant]]
Output: [[VapiAssistant]]

deactivate

Deactivate the assistant.

Input: [[VapiAssistant]]
Output: [[VapiAssistant]]

clone

Clone assistant with all configuration.

Input: [[VapiAssistant]]
Output: [[VapiAssistant]]

test

Initiate a test call with the assistant.

Input: [[TestCallRequest]]
Output: [[Call]]

addTool

Add a tool to the assistant.

Input: [[VapiTool]]
Output: [[VapiAssistant]]

removeTool

Remove a tool from the assistant.

Input: [[VapiTool]]
Output: [[VapiAssistant]]

updatePrompt

Update system prompt.

Input: [[VapiPrompt]]
Output: [[VapiAssistant]]

updateVoice

Change voice configuration.

Input: [[VapiVoice]]
Output: [[VapiAssistant]]

updateModel

Change LLM model configuration.

Input: [[AssistantModel]]
Output: [[VapiAssistant]]

analyzePerformance

Analyze assistant performance metrics.

Input: [[AnalysisRequest]]
Output: [[CallAnalytics]]

exportCalls

Export call data and transcripts.

Input: [[ExportRequest]]
Output: [[URL]]

Events

  • created: Triggered when a new assistant is created.
  • updated: Triggered when assistant configuration is modified.
  • deleted: Triggered when assistant is deleted.
  • activated: Triggered when assistant is activated.
  • deactivated: Triggered when assistant is deactivated.
  • Call.started: Triggered when a call begins with this assistant.
  • Call.ended: Triggered when a call ends.
  • Tool.executed: Triggered when assistant executes a tool.
  • Performance.threshold: Triggered when performance metrics cross thresholds.