Intent Types define the categories of user intentions that the AI Employee can recognize and respond to. Each intent type represents a possible goal or purpose of the user's interaction, such as requesting a service, submitting feedback, updating personal information, or seeking support.
Primary objective
An intent type typically maps to one or more scenarios, guiding the conversation flow based on what the user wants to accomplish. These definitions form the foundation of the AI Employee’s decision-making process, allowing it to route the conversation to the appropriate conversational flow dynamically.
Formatting rules
1. Title
Always use: ## [L/T if applicable] **<Intent Name>**
Examples:
## **General Inquiry**## **[L] Appointment Request**## **[T] Cancellation**
Intent classification markers – optional:
[L] - Lead Intent: Used when the conversation represents a potential lead, opportunity, or conversion (e.g., a request, reservation, or inquiry that could result in a sale or engagement).
[T] - To Do Intent: Used when the conversation outcome requires manual follow-up or internal action by a manager or team member after the session (e.g., callbacks, confirmations, or approvals).
2. Body
Each intent must describe a single, clearly defined user goal or request. If the Intent description matches the user's intent in the conversation, tell the ConvoAgent to perform a specific action. The action can be time-based, for example, during working hours, do this, or during non-working hours, do that. The format is as follows:
[Intent description]
[ConvoAgent/general actions]
Attribute references and context injection
Intents can reference Attributes using the format: [[attribute Idn]]
Example: [[project_representative_agent_name]]
Example
Title
## **[T] Reconfirm Reservation or Order Details**
Body
When the User wants to confirm an existing reservation or has questions about previous or current reservation details, billing, previous food order details, or anything else that ConvoAgent cannot retrieve from the <BusinessContext>.
During Working Hours: ConvoAgent says something like: "As an AI assistant, I'm not able to help you with this request. I'll transfer you to a team member who can help." and starts the scenario: **"Regular Transfer"**.
During Non-Working Hours: ConvoAgent starts the scenario: **"Relaying Message to the Manager"**.