Datablox

Visualizations represent valid SQL queries from the workspace
Increasing difficulty challenges help students learn SQL Queries
Live tables give students the chance to see how their queries look in a terminal
Dark Mode allows for more accessible environments

Inspiration

A future stamped with a growing reliance on technology and data security has created a demand for a skilled workforce starting in our elementary schools. Non-profits across the nation attempt to support the knowledge gap with minimal time, people, and tools.

What it does

Datablox delivers an easy to use workspace that allows all ages the ability to drag and drop SQL block families into structured queries, making database logic intuitive, visual, and immediately executable. Series of goals give students the freedom to experiment with SQL commands using interlocking blocks. Child friendly visualizations bring databases to life and a terminal window displaying structured queries combine to reinforce students confidence during their journey.

How we built it

We used a combination of agentic workflows and LLM supported IDE's to bring our dream to life. Our Built with Next.js, TypeScript, and a modular SQL-game architecture, Datablox combines a Blockly-driven front end with internal parsing, validation, and level systems to teach SQL through structured interaction.

Challenges we ran into

We ran into issues with the logic of SQL terminology. Creating a firm logic base where puzzle blocks can only be logically placed without collisions proved to be the biggest challenge. When trying to implement ElevenLabs with Gemma as the backend LLM we discovered that our underlying infrastructure did not support ElevenLabs SDK and would require a Hugging Face adaptation which we opted not to use for security purposes.

Accomplishments that we're proud of

Creating an interlocking block system that is capable of taking SQL queries that abide by SQL syntax. Creating a live visual representations of SQL queries was also a big accomplishment.

What we learned

We learned how well we understood python ourselves through extensive testing! We also learned that we should anticipate having to use third party dependencies to use certain vendor tools.

What's next for Datablox

Datablox will add a challenge button that creates new challenges using generative AI backed by Gemma 4. We also will add support to grow our tables and introduce SQL commands. We also can not wait to introduce Broncy who will be represented by an on screen horse that acts as a voice assistant that answers students questions using voice and RAG supported agents.

Victor Sah acted as Project Manager, Alexander Okonkwo served as full stack developer, Nicolas Pena took the role of user experience and QA , Jesus Uriarte supported as prompt engineer.