Demo: Generate SQL queries from a JSON Schema
Type your request in natural language — AI-SQL-Chat will build a single, safe SELECT with validation and a limit. You can paste your database schema directly on the Paste Schema page or use a previously generated schema.json (you can generate it in 3 steps).
How it works
On this page, in the field “Ask a question about your data”, write a request in natural language — the AI will convert it into an SQL query. You can either display the generated query or view its result as a table.
The demo database contains data about smart-home modules — including object type (house/apartment), surface area, list of modules (e.g. shutter controller, lighting, RGB), and cities.
show 5 best-selling modules in Warsaw with pricesshow number of sold modules grouped by house/apartment with total pricingshow valuation of modules grouped by home surface area
Ask a question about your data
Write in natural language — we will generate one safe SELECT.Short demonstration
Fill the fields and click “Generate SQL”.Generated SQL
-- Generated SQL query will appear here
Schema preview
Natural language to SQL generator — convert descriptions into SELECT
AI-SQL-Chat acts as a natural-language SQL generator: you type a sentence in English,
and the tool builds a single safe SELECT query based on your schema.json.
Thanks to type validation and column-name matching, the output is ready to run.
Our description → SELECT converter works in read-only mode (no DDL/DML), enforces LIMIT,
and blocks risky SQL constructs.
You can paste your schema directly on the page
Paste schema — just copy the table/column structure
following the short guide at Copy schema guide.
If you need to prepare a schema, see
“How to create schema.json”, then upload it at
“Upload schema”.
This solution speeds up ad-hoc analytics and improves collaboration between business and IT — from sales queries to quick reports and hypothesis testing. See how your description becomes a working SQL query.