AI node - GPT Dialog

Utilize GPT Dialog Node to develop conversational bots. Gather user information through realistic conversations. Control AI interactions using natural language prompts.

Who can use this Card?

• Gallabox Users having permission to either build and deploy bots or manage bots.

• Available on Pro Plan.

Overview

The GPT Dialog Node in Gallabox allows you to create conversational bots that can understand and respond to natural language input. With this card, you can build bots that collect information from users using human-like conversations.

It contains the below sections:

1. Prompt: The prompt is where you provide instructions to your AI. This is like telling your AI what to do in the conversation. You can use simple, natural language instructions to guide the AI in its interactions.

2. User Message: This is the message provided by the user that will trigger the GPT Dialog card. If this card is the first card in your conversation flow, you can use {{intent.text}}to capture the user's initial message. However, in later parts of the bot-flow, you can store the previous card's response in a variable and use it here. This helps the AI understand the context of the conversation.

3. Max Tokens: Max Tokens determine the maximum number of tokens that the GPT-model can use in its response.
4. Temperature: The temperature setting controls the randomness of the AI responses. A higher value, like 0.8, makes the responses more unpredictable and creative. A lower value, such as 0.2, makes the responses more focused and closely related to the prompt.

5. Function: The GPT model will return its output in a structured format via function calls. A function includes the name and parameters.

   • Name: The function name describes the function's purpose. It tells the AI what this function is meant to achieve.

   • Description: The description provides a brief explanation to the AI about what the function does.

   • Store the parameters in the Variable (optional): You can choose to store the data collected (parameters) from users in a variable. This allows you to use the customer responses effectively in the conversation flow, helping your bot remember and utilize the information.

   • Parameters

     • Parameters are the specific data points you want to collect from the users using AI.

     • Name: The name of the parameter is like a title for the data you want to collect. For instance, if you need to gather the customer's business name, you can name it "business_name"

     • Description (optional): The description gives the AI node more context about what this parameter represents. It's an optional explanation of the data you're collecting.

     • Type: The data type specifies what kind of information you want to collect. It can be a string (text), number, or Boolean (true/false).

Example

Let's say you are using the AI node to collect the lead information.

Prompt:

Assume you are a sales associate at ABC company. You collect information about leads to qualify them. You will ask them the following data.

- Business name
- Industry
- Company size
- email ID 

User Message:

{{intent.text}}

Max Tokens:

1000

Temperature:

0

Function:

Name:

Lead information

Description:

Information of the customer collected

Store Parameters in Variable:

Lead_info

• Parameter 1

Name:

business_name

Description:

name of the customer business

Type:

String

• Parameter 2

Name:

industry

Description:

name of the industry

Type:

String

• Parameter 3

Name:

company_size

Description:

number of employee's in the company

Type:

number

• Parameter 4

Name:

email_id

Description:

email address of the customer

Type:

String

Using the Collected Data

Once the user provides the necessary details to qualify as a lead, the information is stored in the "Lead_info" variable for further processing.

If you want to use, for example, the industry information collected from the customer elsewhere in the conversation flow, you can do so using a format like

"{{variables.Lead_info.industry}}."