​

Agent: Building an Intelligent Voice Experience

​

Basic Settings

• Agent Name — A recognizable name for your Agent.
• Voice Type — Select the Agent’s tone (male or female).
• Agent Description — Internal description of the Agent’s purpose or use case.
• Voicemail Message — What the Agent should say if it reaches voicemail.
• Predefined Prompt — Choose a saved prompt from your Prompt Library.
• Custom Prompt — Add or edit a custom prompt directly.

​

Advanced Settings

• Select a Voice — Pick a specific voice from your created voice assets.
• Model Language — Define the language your Agent will use.
• LLM Provider — Choose your preferred large language model provider (e.g., OpenAI).
• Model Name — Options include:
  • GPT-4
  • GPT-4 Turbo
  • GPT-4 Omni
  • GPT-4 Omni Mini
  • GPT-3.5
  • GPT-3.5 Turbo
    Note: Latency and performance may vary by model.
• Speech-to-Text Provider — Select the transcription provider for incoming speech.
• Webhooks — Add one or more webhooks to stream conversation data in real time.
• Actions — Assign specific Actions the Agent can perform (e.g., send SMS, update lead, transfer call).
• LLM Temperature — Adjust response style:
  • Lower = more deterministic.
  • Higher = more creative.
• Conversation Speed — Controls how quickly the Agent interrupts:
  • Lower = slower to interrupt.
  • Higher = faster to respond (may overlap the user).
• Max Idle Check Count — Maximum number of times the Agent checks if the user is still present before ending the call.
• Idle Time (seconds) — Delay before checking if the caller is still there.
• Call Duration (seconds) — Maximum length of the call.

​

Next Steps

• Explore the API: Create Agent
• Learn more: Actions Guide