Metric Moon

• GIF

• 

• 

• 

Inspiration 🚀

I’ve watched colleagues bounce between a SQL IDE, a BI dashboard, and “that one chart” in a slide deck just to answer a single ad-hoc question. It felt like modern analytics still required a translator. Metric Moon grew from my desire to remove that translation layer so anyone could ask their data and see answers instantly.

What it does ✨

• Natural-language to insight: type a question, hit Enter.
  
• AI data agents: generate & run optimized SQL on your database.
  
• Instant visuals: Metric Moon auto-chooses the best visualization and renders it live.

“How many missions reached Mars each decade?”
↓ seconds later ↓
📊 Interactive bar chart + raw table + copyable SQL

How it was built 🛠️

Mission Milestones 🌙

• Users can spin up and connect their data agent in minutes with zero code or infrastructure hassle with the help of SkyAI Agents.
  
• Auto-selected visualizations, Metric Moon chooses the most meaningful chart and its dimensions for every query.
• Clean, modern UI that’s intuitive and fast to navigate.
  

What I learned 📚

I underestimated how much trust users place in seeing the SQL, even non-technical testers opened the “SQL” tab out of curiosity. Transparency breeds confidence. I also learned that few-shot examples plus lightweight tools outperformed monolithic agent chains for this use case.

What’s next for Metric Moon 🌠

• Persistent workspace: save queries, results, and charts for later recall.
• Multi-database connectors.
• Saved dashboards with scheduled re-runs.
• Fine-tuning of the agent to handle window functions and CTEs.
  

Built With

• amazon-web-services
• bolt.new
• css
• html
• javascript
• react
• skysql
• typescript