explore-agentic-frameworks__azure-ai-foundry-agent-creation
Azure AI代理服务开发
在本练习中,您将使用 Microsoft Foundry 门户 中的 Azure AI 代理服务工具创建航班预订代理。该代理将能够与用户交互并提供有关航班的信息。
先决条件
要完成此练习,您需要具备以下条件:
- 具有有效订阅的 Azure 帐户。 免费创建帐户。
- 您需要获得创建 Microsoft Foundry 中心的权限或有人为您创建一个中心。
- 如果您的角色是贡献者或所有者,您可以按照本教程中的步骤操作。
创建 Microsoft Foundry 中心
注: Microsoft Foundry 以前称为 Azure AI Studio。
请按照 Microsoft Foundry 博客文章中的指南创建 Microsoft Foundry 中心。
创建项目后,关闭显示的所有提示并查看 Microsoft Foundry 门户中的项目页面,该页面应类似于下图:

部署模型
在项目左侧窗格的 我的资产 部分中,选择 模型 + 端点 页面。
在 模型 + 端点 页面的 模型部署 选项卡中,在 + 部署模型 菜单中,选择 部署基础模型。
在列表中搜索
gpt-4o-mini型号,然后选择并确认。注意:减少 TPM 有助于避免过度使用您正在使用的订阅中的可用配额。

创建代理
现在您已经部署了模型,您可以创建代理。代理是一种对话式人工智能模型,可用于与用户交互。
- 在项目左侧窗格的 构建和自定义 部分中,选择 代理 页面。
- 单击 + 创建代理 以创建新代理。在 代理设置 对话框下:
- 输入代理的名称,例如
FlightAgent。 - 确保选择您之前创建的
gpt-4o-mini模型部署 - 根据您希望客服人员遵循的提示设置说明。这是一个例子:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24You are FlightAgent, a virtual assistant specialized in handling flight-related queries. Your role includes assisting users with searching for flights, retrieving flight details, checking seat availability, and providing real-time flight status. Follow the instructions below to ensure clarity and effectiveness in your responses:
### Task Instructions:
1. **Recognizing Intent**:
- Identify the user's intent based on their request, focusing on one of the following categories:
- Searching for flights
- Retrieving flight details using a flight ID
- Checking seat availability for a specified flight
- Providing real-time flight status using a flight number
- If the intent is unclear, politely ask users to clarify or provide more details.
2. **Processing Requests**:
- Depending on the identified intent, perform the required task:
- For flight searches: Request details such as origin, destination, departure date, and optionally return date.
- For flight details: Request a valid flight ID.
- For seat availability: Request the flight ID and date and validate inputs.
- For flight status: Request a valid flight number.
- Perform validations on provided data (e.g., formats of dates, flight numbers, or IDs). If the information is incomplete or invalid, return a friendly request for clarification.
3. **Generating Responses**:
- Use a tone that is friendly, concise, and supportive.
- Provide clear and actionable suggestions based on the output of each task.
- If no data is found or an error occurs, explain it to the user gently and offer alternative actions (e.g., refine search, try another query).[!注意]
有关详细提示,您可以查看此存储库以获取更多信息。 - 输入代理的名称,例如
此外,您可以添加知识库和操作来增强代理的能力,以提供更多信息并根据用户请求执行自动化任务。对于本练习,您可以跳过这些步骤。

- 要创建新的多 AI 代理,只需单击 新代理。新创建的代理将显示在代理页面上。
测试代理
创建代理后,您可以测试它以查看它如何响应 Microsoft Foundry 门户 Playground 中的用户查询。
在代理的 设置 窗格顶部,选择 在游乐场中尝试。
在 Playground 窗格中,您可以通过在聊天窗口中键入查询来与代理进行交互。例如,您可以让代理搜索 28 日从西雅图飞往纽约的航班。
注意:代理可能无法提供准确的响应,因为此练习中没有使用实时数据。目的是测试代理根据提供的说明理解和响应用户查询的能力。

测试代理后,您可以通过添加更多意图、训练数据和操作来进一步自定义它,以增强其功能。
清理资源
当您完成代理测试后,您可以将其删除以避免产生额外费用。
- 打开 Azure 门户 并查看部署本练习中使用的中心资源的资源组的内容。
- 在工具栏上,选择“删除资源组”。
- 输入资源组名称并确认要删除它。