本文是《Headless Use Case: Use a Conversation to Build an Agent | 用自然语言对话构建代理》的练习操作步骤。阅读原文请见：Headless Use Case: Use a Conversation to Build an Agent | 用自然语言对话构建代理。
> Note: This badge uses a beta headless builder experience. Go through it in English.
1. Build an Agent Using Natural Language
- In the chat box labeled "Describe the help agent you want to build", enter:
What kind of service agents can you build?
- Enter:
What kind of service agents are good to build for a hospitality company? - Enter:
Good suggestions, let's go with an agent named 'Bookings Help Agent', and set the agent API name to 'bookings_help_agent'. Add a subagent Reservation Agent, named 'Reservation Management' with actions 'FinalizeReservation' and 'GetReservationTimeSlots'. Then add another subagent Guest Complaint Agent, named 'Guest Complaints' with action 'CreateCase'. - If asked about personality, enter:
Go with 'warm' and 'gracious'. - Enter:
Deploy it. - Select the View in Agents list link.
2. Check the New Agent
- Select Launch to open the environment.
- Open the App Launcher and select Agentforce Studio.
- Select bookings_help_agent.
- Select Subagents and review Reservation Management (FinalizeReservation, GetReservationTimeSlots) and Guest Complaints (Create Case).
3. Preview the New Agent
- In the Builder, select Preview.
- Enter:
What times do you have for the fine dining restaurant tomorrow? - Enter:
Will you book the latest one for me? - Enter:
My name is Sofia Rodriguez. My email is sofia@example.com. - Enter
2for the number of guests, thenYesto confirm.