How to Evaluate AI-Generated SQL on Databricks

An AI assistant can produce a query that runs and returns a plausible result. But how do we know if it's correct? Suppose a user asks: Which enterprise customers had revenue growth last quarter? An