GreenGrid AI Scheduler: Optimize Tests for Clean Energy

• 

  main dashboard

Inspiration

AI research consumes massive amounts of power. We wanted to make innovation sustainable by aligning AI workloads with times when renewable energy is abundant, turning computation into a climate-friendly activity.

What it does

GreenGrid AI Scheduler monitors live grid data to identify when renewable sources—like wind, solar, or hydro—dominate the energy mix. It automatically schedules LLM training or testing during those green windows, reducing carbon footprint without sacrificing performance.

How we built it

We integrated real-time grid APIs with a custom scheduler that syncs task queues to green energy periods. The backend uses Python and Node.js, while the dashboard was built with React and Tailwind for clear visualization of grid “greenness” and task timing

Challenges we ran into

Finding consistent, real-time renewable energy data across regions was tricky. We also had to optimize scheduling so that tasks didn’t pile up when green periods were short or unpredictable.

Accomplishments that we're proud of

We built a working prototype that actually shifts compute loads based on renewable availability—and visualizes the environmental impact in real time. Seeing the carbon savings accumulate felt amazing.

What we learned

Energy-aware computing is both technically challenging and urgently necessary. We discovered that even small scheduling changes can significantly reduce emissions from AI workloads.

What's next for GreenGrid AI Scheduler: Optimize Tests for Clean Energy

We plan to integrate with popular ML frameworks and cloud providers, add predictive modeling for upcoming green windows, and let organizations report verified carbon savings from their AI operations.

Built With

• flask
• matplot
• python
• streamlit