Creating Models from Table¶

Creating a model from a table is a quick way to generate a semantic layer model with auto-detected dimensions and measures. Actian AI Analyst analyzes your table structure and sample data to classify columns automatically.

When to Use This Method¶

Use "From Table" when you want:

Quick model generation from an existing table
Auto-detected columns classified as dimensions or measures
A starting point that you can refine manually
Speed over customization - get a working model fast

How to Create a Model from Table¶

Navigate to Models in the left sidebar
Click Create Model
Select From Table
Choose a data source from the dropdown
Search or browse to find your table
Click on the table to start generation

Table Selection¶

The table list shows:

Table name with schema path (e.g., public.orders)
Existing model count - badge showing if models already exist for this table

Use the search box to filter tables by name.

Generation Process¶

After selecting a table, Actian AI Analyst:

Analyzes columns - Reads column names, types, and sample data
Classifies columns - Determines which are dimensions vs measures
Generates expressions - Creates appropriate SQL expressions
Sets the grain - Identifies the primary key or unique identifier
Detects relationships - Finds foreign key connections

What Gets Generated¶

Dimensions are created for columns that represent:

Dates and timestamps
Categories and enums
IDs and foreign keys
Text attributes

Measures are created for columns that represent:

Numeric values suitable for aggregation
Counts and quantities

Relationships are detected based on:

Column naming patterns (e.g., customer_id → customers)
Foreign key constraints

Tips¶

Review auto-generated items - The AI does its best but may misclassify some columns
Add descriptions - Help agents understand what each dimension/measure represents
Set up filters - Add common filters like "active only" or "last 90 days"
Check relationships - Verify detected relationships are correct