Dataplay

Inspiration

Working with data transformations is often slow and requires technical expertise. We wanted to simplify this by letting users connect to any database, explore its schema, and apply transformations using AI—without writing complex SQL.

What it does

Dataplay connects to a Postgres database, reads the schema and sample data, and visualizes the structure. Users can describe desired transformations in plain English, and the app generates SQL code to execute them. It even updates the schema diagram automatically after changes.

How we built it

We used Superdev as our foundation for UI and frontend logic. The backend integrates with a Postgres database, uses OpenAI’s models for code generation, and renders diagrams dynamically. Each action (connect, transform, visualize) is supported by dedicated agents that collaborate under the hood.

Challenges we ran into

Superdev's backend is currently closed-source, which made it hard to modify logic for custom database operations. We had to work around this limitation and adapt some functionality at the frontend/API level.

Accomplishments that we're proud of

We successfully built an AI-powered transformation engine with a connection and automatic schema visualization. The seamless feedback loop from prompt to execution to diagram is something we’re proud of.

What we learned

We learned how to work around platform constraints, build multi-agent workflows, and integrate real-time AI transformations into existing database tools with minimal friction.