Build powerful chatbots in minutes with our intuitive node-based editor

Dialog node

You can set the text returned to the user,
and you can choose whether to wait for user input after the node returns the text,
or directly run the next node.

Large language model chat node

Use a large language model to communicate with users, supporting OpenAI, Ollama, and HuggingFace local models.
If the model returns a timeout, you can also set another reply to return to the user.
Detail

Conditions node

In this node,
you can determine whether the user input is equal to or contains certain text.
You can also determine the intention of the user's input,
or determine whether the value in the variable is what you expect.

Collect node

Using this node,
you can save all or part of the content input by the user into a variable. Later,
the content can be displayed as user text, conditionally judged,
or submitted to a third-party system through HTTP.

Goto node

Using this node,
you can control the direction of the dialog.
You can jump to another conversation, or goto external link.

External HTTP node

Using this node,
you can request an external HTTP interface,
and you can use this node to send input to the outside.
If you need to get data from external HTTP,
please create a variable and choose the value of the variable to be an HTTP interface.

Send email node

This node can send preset email content or dynamic email content to multiple recipients.
It can send emails synchronously or asynchronously.

The end node

This is an endpoint, indicating that a conversation flow has ended or terminated.

• #1 Prerequisite

  Download prebuild files

  • Docker: https://hub.docker.com/repository/docker/dialogflowchatbot/demo
  • Release page: https://github.com/dialogflowchatbot/dialogflow/releases

  Alternatively, you can download source code and build by yourself

  • Frontend: https://github.com/dialogflowchatbot/dialogflow-frontend
  • Backend: https://github.com/dialogflowchatbot/dialogflow-backend
• #2 Guiding page

  Run the program and use a browser to open the access address output by the program
  On the "Guide" page, click on the button in the image below to go to the main flows management page.

  

  One dialog flow contains many main flows which contains many sub flows.
• #3 Main flow

  Create a main flow if there's none.

  

  Here give a name to main flow, you can change the name anytime.
• #4 Dialog flow

  Create dialog flow

  
  

  On the left, you can add sub-flow.
• #5 Editing

  Edit dialog flow

  
  

  You can drag and drop flow node. Double-click on one node, will popup detail setting.
• #6 Testing

  Test dialog flow

  1. Click 'Publish all sub-flows' button (Will save current sub-flow automatically)
  2. Click 'Test dialog test' button, testing window will popup on the right side
  3. Test it

  
  

• #7 Integrating

  Integrate to your application

  Checkout requet API doc

It's similar to Google's Dialogflow, but with an additional canvas for editing processes. It's also similar to Typebot, but it includes a full answering backend.

It has a feature called flow canvas that allows you to visually edit a conversation flow, test it, and finally publish it to the public.

Currently, it can return discourse to the user and capture user input, and can also execute different nodes through conditional judgment ., See demos.

It is very small. The entire software, including front-end and back-end, is less than 6M in size, very easy to distribute and deploy.

You can modify the listening port of the software so that you can run multiple instances on the same server at the same time to handle different user scenarios.

Once it is downloaded, it can be used directly without installing any dependencies. And the data is stored locally, which can protect the data camp privacy.

Advantages

Easy to use

Simple and intuitive.
Everybody can use it
Just few drag drop and clicks
A dialog flow can then be mapped out

Try Demos:

1. Notification of loan repayment
2. Information collection
3. Notification

Tiny fast and portable

Only ONE executable file (database is generated automatically)
pretty easy for deployment
Relying on AoT compilation technology
Program provides high concurrency and blazingly fast responses

Deal with various issues

Use different combinations of flow nodes
Meet the business requirements of different scenarios
Solve problems encountered by different groups of people

Compatibilities

Front-end support for Firefox, Chrome, Microsoft Edge, Safari, Opera
and other major browsers

The application supports deployment on Linux, Windows Server, macOS Server
and other operating systems.

Easy to integrate

Provides a response interface based on the HTTP protocol

FreeSwitch can also be integrated to enable intelligent speech robots

View docs