Response node
A response node defines the format of pass-through messages sent to users and supports a wide range of message types.
The response node lets you edit content in plain text, such as Chinese and English text with punctuation, or in JSON format. The robot client that interacts with the user is responsible for rendering the final message.
Response content
In the response content, you can reference global variables from the conversation using the following formats:
To reference the normalized value of a global variable: `${IntentName.VariableName}`
To reference the original value of a global variable: `${IntentName.VariableName.origin}`
You can add internal global variables using the Global Variables feature on the left side of the canvas.
To reference an external global variable, use the following format: `${VENDOR_PARAM.VariableName}`
You can enable conditional replies to send different responses based on specific conditions. The system evaluates conditions from top to bottom. When a condition is met, the corresponding response is sent. If no conditions are met, the default response at the top is sent. For more information about the evaluation logic, see Judgement node.
Conversation flow configuration
Exceptions in the conversation flow include timeouts and recognition failures.
Timeout: The action taken when a user does not respond within a specific period.
Recognition failure: The action taken when the robot cannot understand what the user says.
A conversation flow jump specifies the text that the robot sends when the conversation flow returns to this node.
The completion flag indicates that the conversation flow is complete when the flow reaches this node. This flag is used to calculate the completion rate.
Variable configuration
Variables can be used to store results from the flow or for internal logic evaluation.
When you assign a value, you can reference a variable or use a static field. For certain business labels, select the checkbox to display the label in reports.
Voice configuration
If you use a voice robot, such as a voice navigator robot or a voice outbound robot, you must configure these settings for the corresponding service.
Barge-in lets users interrupt the robot while it is speaking.
The silence detection setting determines the duration of silence in milliseconds before the user's audio input is considered a complete sentence.
Keypad input allows the user to respond using the keyboard. The timeout for the voice prompt uses the same timeout setting as custom actions.
Select Send text message to send the voice response content as a text message. You can edit the text message content in the section below.
In the Send to number field, enter a fixed phone number or use `$+SystemVariable` to specify a number.
In the Text message content field, enter the message to send. Use `$+SystemVariable` to include variables in the message.
SSML synthesis uses Speech Synthesis Markup Language (SSML) to control the rhythm and pronunciation of the synthesized voice. For more information, see Introduction to SSML.
API integration
If the robot is integrated using an API, you can configure which parameter variables are passed back.