Ask AI
Skip to main content

Extract JSON from llm response

Function: Extract JSON from AI Response​

This action helps you automatically find and extract structured data (JSON) from a text response generated by an Artificial Intelligence (AI) model. AI models often provide information in natural language, but sometimes you need specific data in a structured format to use in other parts of your application. This action intelligently searches the AI's text for JSON and makes it available for you to use.

Input,

  • LLM Response: A piece of text. This is the full response you received from an AI model. This input is required.

Output,

  • Result: A piece of text. This is the extracted JSON content found within the AI's response. By default, this will be stored in a variable named RESULT, but you can choose a different variable name if needed.

Execution Flow,

Real-Life Examples,

Here are some examples of how you can use the "Extract JSON from AI Response" action:

  1. Extracting User Profile Data:

    • Scenario: You ask an AI to generate a new user profile based on some inputs, and the AI responds with a mix of natural language and structured data.
    • Inputs:
      • LLM Response: "Here is the new user profile you requested: json\{\"username\": \"jane.doe\", \"email\": \"[email protected]\", \"status\": \"active\"\}. Please confirm this is correct."
    • Result: The variable RESULT will contain the text \{"username": "jane.doe", "email": "[email protected]", "status": "active"\}. You can then use this structured data to create a new user record in your database.

  2. Parsing a List of Product Recommendations:

    • Scenario: You ask an AI to recommend products based on a customer's preferences, and the AI provides a list of products in a JSON array.
    • Inputs:
      • LLM Response: "Based on your preferences, I recommend the following products: [{"id": 101, "name": "Wireless Headphones", "price": 99.99}, {"id": 102, "name": "Smartwatch", "price": 149.99}]. Let me know if you'd like more details."
    • Result: The variable RESULT will contain the text [\{"id": 101, "name": "Wireless Headphones", "price": 99.99\}, \{"id": 102, "name": "Smartwatch", "price": 149.99\}]. This list can then be displayed to the user or used to update a product catalog.

  3. Retrieving Configuration Settings:

    • Scenario: An AI provides configuration settings for a new feature, embedded within a longer explanation.
    • Inputs:
      • LLM Response: "To enable the new analytics dashboard, please use the following settings: {"dashboard_enabled": true, "data_source": "google_analytics", "refresh_interval_minutes": 60}. This will ensure all data is up-to-date."
    • Result: The variable RESULT will contain the text \{"dashboard_enabled": true, "data_source": "google_analytics", "refresh_interval_minutes": 60\}. These settings can then be applied to configure the new feature in your application.