Skip to main content

Multi-Turn Conversation Node

The Multi-Turn Conversation node enables sophisticated conversational AI experiences with support for multiple content types, custom instructions, exit conditions, and messaging elements. This node is ideal for creating interactive chatbots, tutoring systems, and dynamic dialogue flows with full control over conversation behavior.

Multi-Turn Conversation node

Basic Usage​

Use the Text, Display Text, and Multi-Turn Conversation nodes for your process to create rich conversational experiences.

Inputs​

The Multi-Turn Conversation node accepts multiple inputs to customize the conversation:

Question​

  • Type: Text for questions or prompts to the user
  • Mandatory: Optional
  • Works best with: Text node, AI General Prompt

Provide questions or prompts that guide the conversation flow.

Reference​

  • Type: Reference materials or context for the conversation
  • Mandatory: Optional
  • Works best with: Text node, Knowledge Bank, File Upload

Supply reference materials or background information for the AI to use during the conversation.

Objective​

  • Type: Text defining the conversation goal or objective
  • Mandatory: Optional
  • Works best with: Text node

Define the overall objective or purpose of the conversation to guide the AI's responses.

Summary​

  • Type: Conversation summary or context
  • Mandatory: Optional
  • Works best with: Text node, AI General Prompt

Provide a summary of previous interactions or context to maintain conversation continuity.

Append Instruction​

  • Type: Additional instructions added to the conversation
  • Mandatory: Optional
  • Works best with: Text node

Add supplementary instructions without overwriting the main instructions.

Overwrite Instruction​

  • Type: Instructions that completely replace default behavior
  • Mandatory: Optional
  • Works best with: Text node

Replace the entire instruction set with custom directives for complete control.

Outputs​

Full Chat Log​

  • Type: Complete conversation history with timestamps
  • Works best with: Display Text, Document Download, AI General Prompt

Captures the entire conversation between the user and the AI, including all messages, inputs, and outputs.

Configuration​

Process Type​

Select the AI model to power your conversation:

  • GPT-4o-mini: Fast and cost-effective for most conversational tasks
  • GPT-4o: Advanced reasoning for complex conversations
  • GPT-3.5: Efficient model for straightforward dialogues
  • Gemini 1.5 Pro: Google's advanced conversational model
  • Claude models: Anthropic's AI with strong reasoning
  • Other supported models: Additional AI models available

Content​

Configure what content the conversation should include:

Question (Toggle)​

Enable or disable question handling in the conversation.

Reference (Toggle)​

Enable or disable reference material usage during the conversation.

Instructions​

Append Instruction​

Add supplementary instructions that enhance the default behavior without replacing it.

Overwrite Instruction​

Replace all default instructions with your custom directives for complete control over conversation behavior.

Messaging Elements​

The Bot Name​

  • Type: Text input (green dot)
  • Purpose: Define the chatbot's display name
  • Works best with: Text node

Set a custom name for your conversational AI that users will see.

Icon Image​

  • Type: Image input (pink dot)
  • Purpose: Display a custom avatar for the bot
  • Works best with: Image node, File Upload

Upload an icon or avatar to represent your bot visually.

Placeholder​

  • Type: Text input (green dot)
  • Purpose: Placeholder text in the user input field
  • Works best with: Text node

Define helpful placeholder text that appears in the user's message input area.

First User Message​

  • Type: Text input (green dot)
  • Purpose: Pre-populate the initial user message
  • Works best with: Text node

Set the first message from the user to start the conversation automatically.

Exit Criteria​

Configure conditions for ending the conversation:

Exit Condition for Pass (Optional)​

  • Type: Text input (green dot)
  • Purpose: Define successful completion criteria
  • Example: "User correctly answered all questions"

Specify conditions that indicate successful conversation completion.

Exit Condition for Fail (Optional)​

  • Type: Text input (green dot)
  • Purpose: Define failure or exit criteria
  • Example: "User requested to exit or provided incorrect answers"

Specify conditions that trigger conversation termination due to failure or user request.

Temperature Control​

Temperature Slider​

Control the creativity and randomness of AI responses:

  • Lower values (0.0-0.3): More focused, consistent, and deterministic responses
  • Medium values (0.4-0.7): Balanced between consistency and creativity
  • Higher values (0.8-1.0): More creative, varied, and exploratory responses

Adjust the slider to find the right balance for your conversation style.

User Interaction Options​

Math Input​

  • Type: Checkbox
  • Purpose: Enable mathematical expression input and rendering
  • Use case: When conversations involve equations or mathematical notation

Handwriting Input​

  • Type: Checkbox
  • Purpose: Enable handwriting recognition for user input
  • Use case: For touch/stylus devices or handwritten responses

Attach File Input​

  • Type: Checkbox
  • Purpose: Allow users to upload files during the conversation
  • Use case: When users need to share documents, images, or other files

Voice Typing​

  • Type: Checkbox
  • Purpose: Enable voice-to-text input for users
  • Use case: For hands-free input or accessibility

Write To Journal​

  • Type: Checkbox
  • Purpose: Automatically save conversation to journal
  • Use case: For record-keeping or later analysis

Set Character Limit​

  • Type: Checkbox
  • Purpose: Limit the number of characters in user responses
  • Use case: Enforce concise responses or manage input length

Example Workflows​

Biology Cell Structure Tutoring​

Scenario: Create an AI tutor that helps students learn about cell structures with guided questions and contextual references.

Multi-Turn Conversation Example

Steps to Create the Flow:

  1. Start with the Start Node.

  2. Add Display Text nodes with instructions:

    • First Display Text: "#### Bloodshot Bloodhsot***"
    • Second Display Text: "#### Dr. Yao figure below shows a red blood cell in a body. what is that?"
    • These provide context and initial prompts

  3. Add Text nodes for conversation configuration:

    i. Question Text node:

    #### (a) Select two structures in the root hair cell which are 
    also present in the red blood cell. In roote,roote,stoate,the...
    • Connect to Question input

    ii. Reference Text node:

    Image Description Begins:
    The image includes two scientific diagrams, one of a...
    • Connect to Reference input

    iii. Objective Text node:

    You are an AI tutor tasked with assessing a student's 
    understanding of cell...
    • Connect to Objective input

  4. Configure the Multi-Turn Conversation Node:

    i. Select Process Type: GPT-4o-mini (for cost-effective tutoring)

    ii. Enable Content options:

    • Check "Question" to include the question in the conversation
    • Check "Reference" to provide reference materials

    iii. Configure Instructions:

    • Add Append Instruction for additional guidance (optional)
    • Or use Overwrite Instruction for complete control

    iv. Set Messaging Elements:

    • Connect Bot Name text node (optional)
    • Connect Icon Image (optional)
    • Connect Placeholder text (e.g., "Type your answer here...")
    • Connect First User Message if you want to pre-populate

    v. Configure Exit Criteria:

    • Pass condition: "If the user has change topic: Exit with fail."
    • Fail condition: "If the user correctly answered the question, Do not provide summary when pass, press when pass,..."

    vi. Adjust Temperature: Set to medium (around 0.5) for balanced tutoring responses

    vii. Enable User Interaction Options:

    • Check "Math Input" for mathematical expressions
    • Check "Attach File Input" if students need to submit images
    • Check "Write To Journal" to save conversation history

  5. Add Display Text nodes to show outputs:

    • Connect Full Chat Log to Display Text for review
    • Users can see the entire conversation history

Result:

Students interact with an AI tutor that:

  • Asks guided questions about cell structures
  • Provides reference materials and context
  • Evaluates student understanding
  • Offers feedback and explanations
  • Maintains conversation flow with clear objectives
  • Exits appropriately based on success/failure criteria

Notes​

  • The conversation maintains context throughout the interaction
  • Multiple input types (text, voice, handwriting, files) enhance accessibility
  • Exit conditions enable automatic conversation completion
  • Temperature settings affect response style and creativity
  • Full chat log captures complete conversation history for analysis
  • Journal writing enables persistent storage of conversations