GreenOps
🌿 Autonomous AI agent that reduces cloud carbon emissions by 40% through intelligent scheduling. Powered by Gemini 1.5 Flash & UK National Grid Carbon Intensity API.
Install / Use
/learn @Bhagirath00/GreenOpsREADME
🌿 GreenOps Enterprise
</div>Overview
GreenOps is an autonomous AI agent that optimizes cloud deployments based on real-time carbon intensity. Instead of deploying blindly, it checks the grid status and intelligently schedules workloads for "green windows"—reducing carbon emissions by up to 40%.
Powered By
- Google Gemini 1.5 Flash: The reasoning engine that decides when to deploy.
- UK National Grid Carbon Intensity API: Real-time energy generation data (Wind, Solar, Gas, Coal).
Want to Run Locally?
A dedicated local execution mode for testing on your laptop is here!!
Click Here for Local Setup Guide
The Problem vs. Solution
The Problem: Invisible Carbon
- Cloud computing emits more carbon than the airline industry.
- Deploying a heavy job at 6 PM (Coal/Gas) = 300g CO2/kWh.
- Deploying the same job at 2 AM (Wind) = <50g CO2/kWh.
- Most DevOps tools ignore this completely.
The Solution: Autonomous Agent
GreenOps acts as a "Sustainability Gatekeeper":
- Intercepts deployment requests.
- Checks live grid data via API.
- Decides:
- ✅ Green Grid? Deploy immediately.
- 🏭 Dirty Grid? Schedule for later (e.g., 3 AM).
- 🚨 Critical Emergency? Override and deploy instantly.
Architecture
<img width="1112" height="567" alt="Image" src="https://github.com/user-attachments/assets/3149f53b-72d1-4f96-ab9e-e28c93ba7d2e" />Process Flow
<img width="2724" height="423" alt="Image" src="https://github.com/user-attachments/assets/a103340d-71bd-44d4-a0d4-5f938ad2b371" />Real-World Scenarios
| Scenario | Grid Status | Agent Decision | Impact | |:---|:---|:---|:---| | Nightly Build | Dirty (Gas) | Wait 6 hours for Wind peak | -84% Carbon | | Security Hotfix | Dirty (Gas) | Deploy Immediately (Critical) | Safety First | | Data Migration | Moderate | Wait 22 hours for Solar peak | Human Choice |
Acknowledgments
Built for the Google AI Agents Hackathon (November 2025).
- LLM: Gemini 1.5 Flash
- Data: Carbon Intensity API (National Grid ESO)
Future Improvement
We are building the future of sustainable DevOps. Here is what's coming next:
- [ ] Kubernetes Operator: A K8s controller to auto-schedule pods based on node carbon intensity.
- [ ] Multi-Region Arbitrage: "Follow the Wind" routing—moving workloads from UK (calm) to Spain (windy) automatically.
- [ ] Hardware Awareness: Integration with NVIDIA MIG to throttle GPU power during dirty grid windows.
- [ ] CI/CD Plugins: Native plugins for GitHub Actions and GitLab CI.
🤝 How to Contribute
We welcome contributions from the community! Whether it's fixing bugs, adding new features, or improving documentation.
- Fork the repository.
- Create a Branch:
git checkout -b feature/AmazingFeature. - Commit your changes:
git commit -m 'Add some AmazingFeature'. - Push to the branch:
git push origin feature/AmazingFeature. - Open a Pull Request.
Ideas for Contribution
- Add support for other grid APIs (e.g., ElectricityMaps, WattTime).
- Create a Docker container for the agent.
- Improve the visualization dashboard.
License
Distributed under the License See for more information.
<div align="center">
Code Smarter. Breathe Easier. 🌿
</div>