Configuring SmartSpace.ai

Guide to SmartSpace.ai setup: Sync data, integrate Azure, manage connectors, and configure workspaces for streamlined AI deployment and data analysis.

Configuration – Output Schema

User Guide: Configuring Optional Output Schema in SmartSpace.ai

SmartSpace.ai
offers the option to configure an output schema within workspaces, allowing users to define the structure of the response generated by the workspace. This guide will provide instructions on how to set up and use the optional output schema feature.

Accessing Optional Output Schema

Step 1: Navigate to Workspace

  • Access the desired workspace where you want to configure the output schema.

Step 2: Locate Basic Information Tab

  • Within the workspace, navigate to the “Basic Information” tab.

Step 3: Access Output Schema Section

  • Look for the section related to the output schema. It should include a text box where users can input their desired JSON schema.

Guidance on JSON Schema Creation

Step 1: Define JSON Schema

  • You can create any valid JSON schema, including objects, arrays, numbers, and strings, to structure the output response according to your requirements.

Step 2: Format Schema

  • Ensure that your JSON schema is correctly formatted. You can use online resources or tools to assist with schema creation.

Step 3: Prettify Schema

  • Utilize the “Prettify” button provided within the workspace to format the entered schema f

Schema Enforcement

  • The workspace will adhere to the specified JSON schema, structuring the output response accordingly.
  • It’s advisable to include a pre-prompt indicating the desired output, such as ‘Return a list of products and their corresponding information,’ to ensure clarity in the generated response.

Example Schema

Here’s an example schema for reference:

{
  "type": "object",
  "properties": {
    "title": {
      "type": "string",
      "description": "Title of the document"
    },
    "contents": {
      "type": "array",
      "description": "List of contents for the document",
      "items": {
        "type": "object",
        "properties": {
          "headerName": {
            "type": "string",
            "description": "The name of the main header"
          },
          "subHeaders": {
            "type": "array",
            "description": "Array of names for the subheaders",
            "items": {
              "type": "string",
              "description": "The name of a subheader"
            }
          }
        },
        "description": "An object representing a section in the document with a main header and subheaders"
      }
    }
  }
}

Result from the Workspace

After configuring the output schema, the workspace will generate responses structured according to the defined schema. Here’s an example result based on the provided schema:

{
  "title": "Company's Data Usage Policy",
  "contents": [
    {
      "headerName": "Introduction",
      "subHeaders": [
        "Purpose of the Policy",
        "Scope of the Policy"
      ]
    },
    {
      "headerName": "Data Collection",
      "subHeaders": [
        "Types of Data Collected",
        "Methods of Data Collection"
      ]
    },
    {
      "headerName": "Data Usage",
      "subHeaders": [
        "How We Use Your Data",
        "Legal Basis for Data Usage"
      ]
    }
  ]
}

By following these steps and utilizing the optional output schema feature, you can customize the structure of the response generated by SmartSpace.ai workspaces to meet your specific needs.

keyboard_arrow_up