Creating a New Decode Table

Creating a Table is the first step towards organizing and analyzing your insurance data within Decode. This process involves defining the structure (schema) that will hold your cleaned data. You'll use a multi-step form to specify the different types of columns your table will contain.

Accessing the Create Table Form

1. Navigate to the Data Page.
2. Locate and click the "Create New Table" button (displayed as a plus icon at the top of the left sidebar). This will open the multi-step table creation form.

Step 1: Basic Information

This step captures the fundamental details about your table.

Basic Info

• Table Name: Enter a descriptive name for your table (e.g., "US Property Premiums", "Commercial Auto Claims"). This is the name you'll see in the Decode interface. Spaces are allowed here.
  • Note: Decode will automatically create a system name by converting spaces to underscores and removing special characters.
• Description: Provide an explanation of what data this table holds. This helps you, your team, and Decode understand its purpose later (e.g., "Monthly premium bordereaux data for US property policies underwritten via Wholesaler X"). Detail about the data and how it will be stored is important here as it provides more context for the Decode Agent and function building AI.
• (Optional) Import Template: This allows you to import a predefined template for your data. This template can be edited to match your needs, but provides a strong starting point.

Click "Next" to proceed.

Step 2: Column Definition

This is where you define the core structure: the Row Identifier, Metadata columns, and Static columns.

Setup Columns

Row Identifier

This is the most important column you will define.

• What it is: The single column that uniquely identifies each row in your table (e.g., each distinct policy or claim).
• Why it matters: Decode uses this identifier to track records over time. When you upload new data, Decode checks this column to see if a row represents an update to an existing record or if it's a brand new record.
• Fields:
  • Column Name: Enter the exact name of the column in your typical spreadsheets that contains this unique identifier (e.g., Policy Number, Claim ID). Crucially, this column MUST exist in the files you plan to upload.

Metadata Columns

These columns allow you to add contextual information or tags during the upload process. The data for these columns does not need to be present in the uploaded spreadsheet itself.

• Fields:
  • Column Name: Select an existing metadata category from the dropdown.
  • Description (Optional): Explain what this tag represents.
  • Mandatory: Check this box if a value must be provided for this tag during every file upload to this table. Uploads will fail if a mandatory metadata tag is missing.
• Adding/Deleting: Use the "Add Metadata Column" button to add more rows and the Delete icon (🗑️) to remove them.

Static Columns

These columns represent attributes of your records that generally don't change from one reporting period to the next.

• Fields:
  • Column Name: The name of the column as it often appears in your spreadsheets (e.g., Insured Name, Policy State, Coverage Type).
  • Description (Optional): Explain the column's content.
  • Mandatory: Check this box if this column must be present and have a value in every uploaded file destined for this table. If data is missing for a mandatory static column in any row of an uploaded file, the file processing will fail.
• Adding/Deleting: Use the "Add Static Column" button and the Delete icon (🗑️) as needed.

Click "Next" when you have defined these columns.

Step 3: Dynamic Columns

This section is for defining data that changes or is reported periodically (e.g., monthly premiums).

Dynamic Column Builder

• Enable Dynamic Columns: Use the toggle switch to activate this feature. If your data doesn't have a periodic component (e.g., it's just a static list), you can leave this disabled.
• Structure Definition: If enabled, you define the template for the data that will be stored for each period. Decode will automatically create new columns based on this template as data for new periods (like months) arrives.
  • Adding Values/Groups: You can define:
    • Single Values: Direct metrics for the period (e.g., Gross Premium Reported).
    • Groups: Collections of related metrics (e.g., a Financials group containing Gross Premium, Commission, Net Premium). This helps organize complex periodic data.
  • Fields (for each value/nested value):
    • Name: The name of the specific metric (e.g., Paid Claims).
    • Update Behavior: How should Decode handle this value if it already exists for a given record and period?
      • Overwrite: Replace the old value with the new one (default).
      • Add: Add the new value to the existing value (useful for cumulative numbers).
      • Subtract: Subtract the new value from the existing value.
    • Mandatory: Must this specific metric be present in the data for every period?
  • Adding/Deleting: Use the "Add Single Value", "Add Group", and delete icons within the builder interface.

Click "Next" to review your schema.

Step 4: Secondary Identifier

This step is simple but important. For each table you must select a secondary identifier. This is used by Decode to prepend to the row identifier if the value for a given spreadsheet is too simple and may cause clashes with other data.

Example:

Policy Number	Contract ID	Other Data ...
01	20021	....
02	20021	....

With the secondary identifier set to Contract ID, Decode will ingest the data by setting the data like so:

Policy Number	Contract ID	Other Data ...
20021-01	20021	....
20021-02	20021	....

in the database. This is crucial for prevent clashes between rows that are not actually the same.

Selecting a Secondary Identifier

Note: Any row identifier less than 5 characters long is considered too simple, and will have the secondary identifier prepended to ensure safety.

Step 5: Review

This final step displays a preview of the complete table schema you have just defined.

Table Preview

• Carefully Review: Double-check all column names, types, mandatory settings, and the dynamic column structure (if enabled). Ensure the Row Identifier is correct.
• Go Back if Needed: Use the "Back" button to navigate to previous steps and make corrections.

Creating the Table

Once you are satisfied with the schema preview:

1. Click the "Create Table" button.
2. Decode will now:
   • Create the corresponding table.
   • Save the table definition (schema, metadata) in our database.
   • Make the new table available on the Data Page sidebar.

Congratulations! Your table is now ready. You can proceed to manage Metadata, create Data Profiles, or start uploading data to it.