SpeakSQL

Data interface · 2025 · Archive 08 / 11

A text and voice interface that turns plain-language questions into SQL and runs them against imported data.

SpeakSQL was built for FutureHack's natural-language data challenge. It lets someone explore a database without knowing the exact query they need to write first.

What I made

Users can upload a CSV or connect PostgreSQL, MySQL, and SQLite. SpeakSQL reads the schema, accepts a typed or spoken question, and returns the generated query before execution.

There is still a direct SQL editor for people who want it. Results appear in a table and can be exported as JSON, CSV, or Excel.

How it works

5 steps
01Connect data

Upload CSV or choose a supported database.

02Read schema

Map tables, columns, types, and relationships.

03Ask

Type or speak a question in ordinary language.

04Generate

Turn the request into visible SQL or Python.

05Run and export

Execute, inspect the table, and save the result.

A compact map of the main product loop.

Repository breakdown

This comes from GitHub's detected language breakdown. It measures source size, not the time or difficulty of the work.

Language mix

GitHub source bytes
TypeScript95%
CSS4%
JavaScript<1%
Role

Hackathon build

Status

Hackathon build

SourcePublic GitHub repository