AI Text2SQL

The example below shows AI generating 5 select queries for a given dataset then explaining what some SQL does:

How does it work?

Behind the scenes QStudio gathers some information on your schemas, your table names, your column names and together with some warm-up questions specific to your database technology, asks openAI to generate more queries:

Running AI Queries

QStudio includes a number of features that enables AI assistance.

These abilities are available from the File Menu, the Toolbar and by right-clicking on a section of code within the editor

In the example, I have:

1. Typed an SQL comment asking OpenAI to generate 5 queries for a given dataset.
2. Pressed F7 or selected "AI Text to SQL".
3. The AI has produced 5 queries which are inserted below your request.
4. You can then run the queries.

Running the Example Yourself

To run the example yourself, you can download a .parquet example file and in windows simply double click it or on other platforms, enter the code below:

CREATE OR REPLACE VIEW titanic AS SELECT * FROM read_parquet('C:\temp\parq\titanic.parquet'); SELECT * FROM titanic LIMIT 111000; -- Press F7 on the comment line below to use AI to generate queries -- Generate 5 example queries for the table titanic. Include a simple select, a count and 3 more advanced queries. SELECT "PassengerId", "Name", "Age" FROM titanic LIMIT 5; -- 2. Count of passengers who survived: SELECT COUNT("PassengerId") as total_survived_passengers FROM titanic WHERE "Survived" = 1; -- 3. Advanced query 1: List all passengers who were in Pclass 1 and survived: SELECT "PassengerId", "Name" FROM titanic WHERE "Pclass" = 1 AND "Survived" = 1 LIMIT 10; -- 4. Advanced query 2: Find the average age of passengers in each class: SELECT "Pclass", AVG("Age") as avg_age FROM titanic GROUP BY "Pclass"; -- 5. Advanced query 3: List the top 3 most common names among passengers: SELECT "Name", COUNT("Name") as name_count FROM titanic GROUP BY "Name" ORDER BY name_count DESC LIMIT 3;

kdb+ AI SQL Generation