Perform SQL Query by Natural Language

Intermediate

Ask questions of table data with generated SQL, DDL scoping, actions, and streaming behavior.

What you'll learn

How Action changes the response returned
How to scope Data Tables to reduce risk and confusion
Why Get SQL and Get DDL are essential test and debugging tools

Perform SQL Query by Natural Language sends a prompt and selected database schema to a model. The step can return a natural language answer, raw data, generated SQL, or the DDL schema that would be sent to the model.

Stuck is a valid status

Need a second brain on this one?

If this lesson just collided with your real schema, script stack, or deadline, book consulting and turn the confusion into a concrete plan.

Book consulting

1/3

Scope the schema deliberately

Only schema for the specified tables is sent. You can choose tables from a list, provide table names by expression, or supply custom DDL if you need strict control.

FileMaker Script

Perform SQL Query by Natural Language [
  Account Name: "main-ai" ;
  Model: Global::ModelName ;
  Prompt: "Which five customers have the highest unpaid invoice balance?" ;
  Action: Query ;
  Data Tables: By name ;
  Tables: "Customers" & Char ( 13 ) & "Invoices" ;
  Response Target: $$Answer
]

Perform Find by Natural Language Embeddings and Perform Semantic Find