Text-to-SQL

Platform · Text-to-SQL

Ask your database in plain English.
Get SQL, results, and the reasoning.

Text-to-SQL is a full data-querying world inside AIRaML. Ask a question the way you'd say it; a free LLM writes SQL for your schema, runs it live against a real database, and explains both the query and the results — its own FastAPI backend behind it.

Open the toolSee how it works
Live databaseMulti-provider LLMExplained queriesDemo · CSV · your DB

You ask

Show me the top 5 artists by total album count.

on the Chinook demo · 11 tables · ~15.6k rows

Text-to-SQL writes

SELECT ar.Name AS artist,
       COUNT(al.AlbumId) AS albums
FROM Artist ar
JOIN Album al
  ON al.ArtistId = ar.ArtistId
GROUP BY ar.ArtistId
ORDER BY albums DESC
LIMIT 5;

A real example. The SQL is AI-generated — the tool shows it and explains it, so you can verify before trusting it.

One workspace, four capabilities

Question in, answer out

Ask → run → explain, over a database you choose. Every capability is live today — the whole pipeline runs against a real database, not a canned preview.

● Live

Ask

Type a question the way you'd say it out loud — "top 5 artists by album count" — and a free LLM turns it into SQL for your schema.

Groq · Mistral · Cohere

● Live

Run

The SQL runs live against a real database, with pagination and column filtering on the results — not a canned preview.

Live query · paginated

● Live

Explain

An LLM explains the generated SQL and the results in plain English, so you can trust what ran and what came back.

Query + results, in words

● Live

Connect

Start on the Chinook demo, upload a CSV, or point it at your own PostgreSQL, MySQL or SQL Server. Browse the schema and its ER diagram.

Demo · CSV · Postgres · MySQL

Ask→Run→Explain→Connectnatural language in, explained SQL out

How it works

Pick → ask → understand

01

Pick a database

Start on the Chinook demo, upload a CSV, or connect your own PostgreSQL / MySQL / SQL Server. Browse the schema and its ER diagram.

02

Ask in plain English

Type a question — "total revenue for each year". A free LLM (Groq → Mistral → Cohere) writes SQL for your schema; you see the query before it runs.

03

Run and understand

The SQL runs live with pagination and filtering, and an LLM explains the query and the results in plain English.

Honest scope

Real queries, with the caveats stated

The SQL is written by a language model and runs against a real database. That is powerful and fast — and it means the output is a draft to check, not an oracle.

LiveWhat it does

  • ✓Natural-language → SQL for your own schema, live
  • ✓Real execution with pagination + column filtering
  • ✓Plain-English explanation of the query and the results
  • ✓Chinook demo, CSV upload, or your Postgres / MySQL / SQL Server
  • ✓Schema explorer + ER diagram
  • ✓Multiple free LLM providers, with fallback

CaveatsWhat to keep in mind

  • —AI-generated SQL — verify results before you rely on them
  • —Aimed at reading / analysing, not bulk unbounded exports
  • —Results are paginated and bounded, not a full data dump
  • —No writes/DDL by intent — it's for asking, not migrating
  • —Your connection details aren't stored server-side

Ask your first question.

Start on the Chinook demo — no setup — or connect your own database.

Open the tool