Re:lation, Dify, and Chatworkを連携して、チャットが開始されたら、回答案をAIで生成しChatworkに通知する方法

"Inquiries are increasing, and we can't finish handling them within business hours." "Even though the team is handling them, the quality of responses varies among staff." Are you facing these issues?
Re:lation is a convenient tool that allows you to centrally manage multiple inquiry channels such as chat and email. By integrating Dify and Chatwork here, you can use it even more conveniently. By utilizing a system where Dify automatically generates response proposals based on inquiry content and notifies them via Chatwork, you can achieve quick and consistent responses.
Such settings, which traditionally required programming, can be easily implemented using our no-code templates. We will explain the setup method and specific benefits in detail, so please read to the end.

Recommended for

This flow bot is recommended for those facing the following challenges.

Those who have many inquiries and tend to exceed business hours in handling them
Those who are concerned about the variation in responses among staff and want to stabilize response quality
Those who find that complaint handling and difficult cases are concentrated on specific staff, leading to uneven burdens
Those who find it troublesome to create response proposals and are troubled by keeping customers waiting
Those who have already implemented Re:lation and Chatwork and are interested in utilizing AI (Dify)

Benefits and specific use cases of generating response proposals with Dify and notifying via Chatwork when a ticket is created in Re:lation

We considered how this automation can be useful from the perspective of customer support representatives, quality control personnel, and customer support team leaders.

1. Secrets to speeding up inquiry responses

The system where Dify quickly generates response proposals and notifies them via Chatwork improves the speed of inquiry responses. When an inquiry ticket is created in Re:lation, Dify generates a response proposal based on past response history and templates. This response proposal is notified to a specific talk room in Chatwork, allowing the person in charge to create a reply based on the response proposal. This saves the trouble of creating from scratch, reducing work time. This automation shortens the waiting time from inquiry to reply. For example, even during a sales period when inquiries are concentrated, AI prepares response proposals, allowing efficient handling. This leads to the elimination of customer waiting times and contributes to service improvement.

2. A system that allows anyone to provide consistent responses

Dify learns company rules and tone to generate response proposals of consistent quality. This prevents variations in responses among staff. By pre-learning company rules and response policies with Dify, unified responses are possible regardless of which staff member handles it. Especially for less experienced staff, they are freed from the burden of research, allowing them to focus on customer interactions. This results in a reassuring response for customers. Also, because responses are consistent, sharing response policies across the team becomes smooth. This system is particularly powerful when it is necessary to advance work while unifying response policies within the team.

3. Preventing personalization and creating a system for team-wide efforts

In customer support, complaint handling and difficult inquiries tend to be biased towards responsible or experienced staff, which is a major factor leading to personalization. By utilizing Dify, AI analyzes customer requests and supports initial responses, allowing any staff member to start responding based on a certain standard. This system allows for team-wide responses without relying on specific staff. For example, even if the inquiry content is difficult, Dify analyzes customer requests and generates standardized response proposals. By notifying these response proposals to a specific talk room in Chatwork, it is shared who is handling which case, and leaders can distribute tasks fairly. Additionally, by utilizing Chatwork's task management function to make it easier to check the progress of responses, delays in progress can be detected early, and appropriate support can be provided. This prevents personalization of responses and establishes a system where roles can be divided and cooperation can be achieved across the team. As a result, it prevents the concentration of burdens on specific staff and improves the productivity and stability of operations for the entire team.

From here, we will explain the setting method for the automation of generating response proposals with Dify and notifying via Chatwork when a ticket is created in Re:lation using Yoom's flow bot template.

[What is Yoom]

How to generate response proposals with Dify and notify via Chatwork when a ticket is created in Re:lation

Utilizing Yoom's trigger function, obtain the ticket information newly created in Re:lation. Based on this obtained information, instructions are sent to Dify, and AI generates response proposals. By notifying these response proposals to stakeholders via the Chatwork API, a system is realized where inquiry information can be shared and handled smoothly. Generally, programming knowledge is required to achieve this, but it is possible without programming knowledge by using the no-code tool Yoom.

Integration process

Register my apps in Yoom (Re:lation, Dify, Chatwork)
Copy the template
Set up app triggers and integrate apps
Turn on the trigger button and start integration

Items to Prepare in Advance

Re:lation

Re:lation account information (for operating the management screen)
Access token

Click "System Settings" at the bottom left of the Re:lation management screen.
Select "API Token" from the menu within System Settings.
If there is an existing value, click the eye icon of the API token to display the value. This value is required.
If you want to generate a new one, you can do so from the "+" button at the top right.

Subdomain
- Can be obtained from the URL. The URL of Re:lation is in the following format:
  "https://{subdomain}.relationapp.jp/"
Inbox
- To specify the ticket inbox, it is smooth if you decide in advance which inbox to use. Re:lation Official: How to Add, Edit, and Delete an Inbox

Dify

Dify account information (for operating the management screen)
App (Chatbot)
- How to create: Dify Official App Creation
Access token
- Obtain from the Dify app.
- Open the created app and select "Monitoring" from the left menu bar.
- Click "API Key" within Backend and Service API.
- Click 'API Key' and select 'Create New Secret Key' to display the API Secret Key. This corresponds to the access token. More details: Dify My App Registration Method and What You Can Do by Linking

Chatwork

Chatwork account information (email address, password)
Room ID to send Chatwork notifications

Help: How to Confirm Chatwork Room ID

Step 1: Setting Up My App Integration

Here, we will set up the registration of the app information to be integrated into Yoom. Log in to the Yoom workspace and click + New Connection on the My Apps screen.

A list of apps that can be connected to Yoom will be displayed. Search for the app to be integrated from the search window and proceed with the registration of My Apps.

Re:lation

Click the app name from the search results. A screen will be displayed to enter the values obtained from Re:lation.

Account Name
- Enter a value that is easy to identify. Use a value that can identify the account, such as an email address.
Access Token
- Enter the API token.

Once you enter the required fields and click "Add," the registration of Re:lation My App is complete.

Dify

Click the app name from the search results. A screen will be displayed to enter the values required for integration.

Account Name
- Enter any value. It is recommended to use a value that can identify the account, such as an email address.
Access Token
- Enter the value obtained in advance.

Once you enter the required fields and click the add button, the registration of Dify My App is complete.

Chatwork

Select Chatwork from the search results. The login screen will be displayed.

Enter the email address and password of the account to be integrated and allow access from Yoom.

This completes the registration of Chatwork My App.

Step 2: Copy Yoom's Flow Bot Template to My Project

Click this banner.

The template will be copied to My Project. You can access My Project from the left menu bar.

Step 3: App Trigger "When a New Ticket is Created"

Click on the first step of the template.

Here, you will set up to retrieve ticket information issued by Re:lation.

(1/2) Select Integration Account and Action

This item is already pre-configured.

App Trigger Title
App
Account Information Integrated with Re:lation
Trigger Action

→ Click Next.

(2/2) API Detailed Settings for App Trigger

Trigger Activation Interval
- Set the activation interval for the Flowbot.
- You can choose from 5 minutes, 10 minutes, 15 minutes, 30 minutes, or 60 minutes. *The shortest activation interval varies depending on your plan.
Subdomain
- Can be obtained from the URL. "https://{subdomain}.relationapp.jp/"
Inbox ID
- Click the edit field to display candidates. When you select the inbox to use, the ID will be quoted.

→ Click Test. → Successful. The following items can be obtained. These are the values used in the next operation.

→ Click Save.

Step 4: Integrate with the App "Retrieve Ticket Details"

Click on the second step of the template. Here, you will retrieve inquiry details from the ticket ID obtained in the previous step.

(1/2) Select Integration Account and Action

This is already pre-configured.

App Trigger Title
App
Account Information Integrated with Re:lation
Action

→ Click Next.

(2/2) API Connection Settings

Subdomain
- Enter the same value as before. "https://{subdomain}.relationapp.jp/"
Inbox ID
- Click the edit field and select the inbox to use from the displayed candidates. Please select the same inbox as in the previous operation.
Ticket ID
- Click the edit field to display the output obtained in the previous step. Select the ticket ID from it, and the quoted code will be embedded.

→ Click Test. → Successful. Inquiry details from the customer can be obtained and will be used in the next operation.

→ Click Save.

Step 5: Integrate with the App "Send Chat Message"

Click on the third step of the template.

Here, you will set up the prompt (instruction) to be sent to Dify.

(1/2) Select Integration Account and Action

This is already pre-configured.

App Trigger Title
App
Account Information Integrated with Dify
Action

→ Click Next.

(2/2) API Connection Settings

query
- Please quote the inquiry content from the customer.
- Click the edit field to display the output obtained in the previous operation. Create instructions to be read by Dify from this.

user
- Set an arbitrary identifier in alphanumeric characters.
Variable

It is not mandatory, but by utilizing it, you can set prompts and notifications more flexibly and efficiently. For example, dynamic information such as "name" or "date" can be automatically reflected in the prompt, reducing the setup effort and enabling optimal response to users.

→ Click Test. → Successful. Dify's response proposal (response content) can be obtained.

→ Click Save.

Step 6: Integrate with the App "Send Message"

Click on the fourth step of the template.

Here, you will set up the template message to be notified to Chatwork.

(1/2) Select Integration Account and Action

This is already pre-configured.

App Trigger Title
App
Account Information Integrated with Chatwork
Action

→ Click Next.

(2/2) API Connection Settings

Room ID
- Click the edit field to display room candidates. When you select the room to notify, the ID will be quoted. Alternatively, you can obtain the ID by checking the Chatwork room settings.
  Help: How to Check Chatwork Room ID

Message Content

Create the template message to notify.

Click the edit field to embed the values obtained in the previous operation (sending a chat message to Dify). By quoting the subject and body code of Re:lation, you can check the consistency with Dify's response content.

The output obtained in each step is divided into tabs, so please check them thoroughly.

→ Click Test. → Successful

→ Click Save. This concludes the steps.

Step 7: Turn ON the App Trigger

Once all steps are configured, a completion screen will be displayed.

Turn ON the trigger and activate the Flowbot with the toggle at the top left of the template.

The Flowbot setup was easily completed without programming. The template used this time is here.

Make the Most of It! 3 Automation Tips Using Dify and Chatwork

Dify and Chatwork can integrate with other apps, allowing you to generate and notify response proposals for inquiries in the same way, which can speed up customer support.

1. Centralize and Efficiently Manage Large Volumes of Emails with Outlook

In handling business emails with Outlook, Dify generates AI-based response proposals. By using a common response base among multiple staff members, standardization of responses can be achieved. Additionally, sharing notifications via Chatwork supports swift progress in handling inquiries.

2. Quick Response Tailored for Simple Operations with Gmail

Dify generates AI-based response proposals for inquiries received via Gmail. Ideal for small team operations and flexible work styles, it helps reduce the burden of individual responses. Chatwork notifications make it easy to share information among team members.

3. Streamline Freshdesk Support Operations

Dify automatically generates response drafts for tickets created in Freshdesk using AI. Agents can efficiently handle inquiries by utilizing these drafts. Additionally, Chatwork notifications facilitate smooth information sharing within the support team.

Summary

In this article, we explained how to set up automation using Re:lation, Dify, and Chatwork. By implementing this automation, you can expect faster response times to inquiries, consistency in response content, and prevention of personalized handling.
Previously, programming was required for these settings, but with this template, you can easily implement it with no code. Minimize the effort required for setup and start automation immediately. Why not achieve more efficient and stable customer support using this template? You can start your trial from the banner below.

See you again!