平台
特点
代理工作室
快速建立和定制您的代理
自主发动机
使用LLMs 引导对话和任务
知识库
使用自定义知识源训练机器人
表格
存储和管理对话数据
渠道
Whatsapp 徽章
WhatsApp
Instagram 徽章
Instagram
FacebookMessenger 徽标
Messenger
Slack 徽标
Slack
所有频道
集成
Hubspot 徽标
HubSpot
Notion 徽标
Notion
Jira 徽标
Jira
Calendly 徽标
Calendly
所有集成
LLM 提供商
OpenAI 徽标
OpenAI
Anthropic 徽标
Anthropic
Groq 徽标
Groq
拥抱脸庞徽标
Hugging Face
全部LLMs
解决方案
对于
企业
实现关键任务生产工作流程自动化
机构
提供完善的代理服务
开发者
探索用于代理开发的强大 API
客户故事
从成功客户那里了解Botpress 如何改变全球业务。
按行业分类
电子商务
教育
财务
接待服务
所有行业
部门
销售
工程学
产品
ITSM
所有部门
按用例
购物助理
潜在客户生成
员工体验
票务管理
所有用例
资源
必要
Academy
通过精心设计的课程学习建造
图书馆
增强人工智能工作流程的资源
博客
有关Botpress 和人工智能代理的见解和最新情况
建筑物
Discord
与成千上万的同行一起交流思想
文件
全面的指南和参考资料
应用程序接口
与外部系统配合使用的参考资料
LLM 排名
比较模式提供商的性能和成本
视频
教程、演示和产品演示
更新日志
了解Botpress 的最新更新
合作伙伴
成为合作伙伴
加入我们的认证专家网络
聘请专家
与合作伙伴和顾问联系
文件
企业
价格
登录
联系方式注册
返回Hub

Weavel

v3.0.0
安装到工作区
维护者 Weavel
  This integration allows you to connect your Botpress chatbot with Weavel, a powerful analytics platform for conversational AI. By integrating your chatbot with Weavel, your chatbot's conversations will be automatically captured and analyzed with Weavel's AI, and configurable reports will be generated automatically to help you understand your chatbot's performance and user behavior.

You can also set up the integration from Weavel, to set up a scheduled data pull from Botpress to Weavel.

## Prerequisites

To establish this integration, the following is required:

- Authorized access to an existing Weavel project. If you don't have an existing account, you can sign up for free [here](https://weavel.ai).
- A valid Weavel project API key.

After the setup, you should follow the instructions below to ensure all of your chatbot's conversation data is properly captured by Weavel.

You can also follow an interactive guide to set up the integration [here](https://weavel.ai/docs/platform-integrations/botpress#log-data-from-botpress-studio).

## Setup realtime message logging

Before activating the Botpress Weavel Integration, please walk through the following steps:

### First, add the Weavel API key to your bot's configuration.

Add your Weavel project API key to your chatbot's configuration variables. This is required to make it accessible to the hooks you will add in the next step.

![Bot configuration](https://i.imgur.com/sDnhkz4.png)

### Second, add two hooks into your chatbot.

> **Note:** You can add hooks in the "Hooks" section of your chatbot's admin panel.
> ![Hooks section of admin panel](https://i.imgur.com/cNzPNyJ.png)

Create a new hook under the "After Incoming Message" section, and add the following code:

```typescript
try {
  const userMessage = event.preview;
  const incomingMessageId = event.id;
  workflow.incomingMessageId = incomingMessageId;
  await axios.post(
    "https://api.weavel.ai/capture/trace_data",
    {
      trace_id: event.conversationId,
      trace_data_id: incomingMessageId,
      user_id: event.userId,
      role: "user",
      content: userMessage,
      metadata: {
        // Add any additional metadata you want to capture (for all messages) here.
        // To capture metadata for specific messages, use the "Log message metadata" card in the studio.
      },
    },
    {
      headers: {
        Authorization: `Bearer ${event.state.configVariables?.["weavelApiKey"]}`,
      },
    }
  );
} catch (error) {
  console.log(error);
}
```

Next, create another hook under the "Before Outgoing Message" section, and add the following code:

```typescript
try {
  let assistantMessage: string;
  if (outgoingEvent.payload.type == "text") {
    assistantMessage = outgoingEvent.preview;
  } else {
    assistantMessage = `${outgoingEvent.payload.type}: ${JSON.stringify(
      outgoingEvent.payload
    )}`;
  }
  const outgoingMessageId = outgoingEvent.id;
  workflow.outgoingMessageId = outgoingMessageId;
  const data = {
    trace_id: event.conversationId,
    trace_data_id: outgoingMessageId,
    user_id: event.userId,
    role: "assistant",
    content: assistantMessage,
    metadata: {
      // Add any additional metadata you want to capture (for all messages) here.
      // To capture metadata for specific messages, use the "Log message metadata" card in the studio.
    },
  };
  await axios.post("https://api.weavel.ai/capture/trace_data", data, {
    headers: {
      Authorization: `Bearer ${event.state.configVariables?.["weavelApiKey"]}`,
    },
  });
} catch (error) {
  console.log(error);
}
```

## Advanced Usage

### Identify users

You can use this function to identify users in Weavel. This will populate the user's profile in Weavel with the provided user id, name, email, and other details.

### Log track event

You can use this function to log track events to Weavel. This is useful for tracking user interactions and events in your bot, such as button clicks, successful form submissions, and other custom events.

### Log message metadata

Use this function to log metadata for specific messages. This is useful for capturing additional context for specific messages, such as user feedback score, details for AI generated messages (knowledge base, AI model, etc.), etc.

#### Some additional setup is required for this action card to work:

1. Add four workflow variables to your bot: `incomingMessageId`, `outgoingMessageId`, `messageId`, and `metadata`. If you have copied the hooks from the previous section, you will see that these variables are used to store the message ids for incoming and outgoing messages.

2. Just after the card that sends/receives the message where you want to log metadata to, add an **Execute code** action card to set the `messageId` variable. Copy one of the following code snippets into the card:

```typescript
// If you want to log the user's message
workflow.messageId = workflow.incomingMessageId;
```

```typescript
// If you want to log the assistant's message
workflow.messageId = workflow.outgoingMessageId;
```

3. Before the **Log Message Metadata** action card, add an **Execute code** action card to declare the metadata as a JSON string. Refer to the example below:

```typescript
workflow.metadata = JSON.stringify({
  feedbackScore: 5,
  model: "gpt-3.5-turbo",
});
```

## Learn more

With the integration enabled, Weavel will analyze your chatbot's conversations and generate reports to help you understand your chatbot's performance and user behavior. To learn more about Weavel's capabilities and how to use the reports, refer to the [Weavel documentation](https://weavel.ai/docs).

用Botpress

打造令人惊叹的人工智能代理体验。

开始免费使用
箭头图标
了解更多信息,请访问Botpress Academy

利用我们精心挑选的课程、指南和教程,更好更快地构建人工智能代理。

聘请专家

与我们的认证开发人员联系,寻找适合您需求的Botpress 专家。

所有系统运行
SOC 2
认证
GDPR
符合要求
© 2025
平台
价格
代理工作室
自主发动机
知识库
表格
Hub
集成
渠道
LLMs
资源
咨询销售人员
文档
聘请专家
视频
客户故事
API参考
博客
状态
v12 资源
社区
社区支持
成为合作伙伴
成为大使
成为会员
公司
关于
职业机会
新闻中心
法律门户
隐私权
©Botpress 2025