The way your AI agent responds to callers depends on several factors. This article explains how responses are generated and how you can influence them.
How Responses Are Created
When the AI agent receives a question or request, it:
- Understands Intent — Determines what the caller wants
- Retrieves Information — Accesses relevant knowledge
- Considers Context — Remembers the conversation so far
- Applies Personality — Uses your configured tone and style
- Generates Response — Creates a natural reply
Factors Affecting Responses
Personality Settings
Your configured personality shapes responses:
- Traits (Friendly, Professional, etc.) influence word choice
- Tone of Voice (Warm, Direct, etc.) affects delivery style
- Response Style (Concise, Detailed, etc.) determines length
Creativity Setting
The temperature/creativity slider controls variability:
- Lower values — More consistent, predictable responses
- Higher values — More varied, creative responses
Custom Instructions
Any specific guidelines you've added direct the agent's behavior and priorities.
Knowledge Base
The information available to the agent determines what it can answer accurately.
Response Patterns
Greetings
Opening messages follow your configured greeting templates exactly, with variables filled in.
Questions
For questions, the agent:
- Searches the knowledge base
- Formulates an answer
- Adds context if helpful
Booking Requests
For bookings, the agent:
- Checks availability
- Presents options
- Confirms details
- Completes the booking
Unknown Requests
When the agent doesn't have an answer:
- Acknowledges the question
- Offers alternatives
- May suggest calling back or speaking with staff
Improving Responses
If responses aren't what you expect:
- Add Knowledge — Expand your knowledge base
- Adjust Personality — Change traits or tone
- Add Instructions — Provide specific guidance
- Tune Creativity — Adjust the temperature setting