Grid-Deploy

Inspiration

Every second, thousands of developers hit deploy. We obsess over latency. We ruthlessly optimize for cost. But almost nobody thinks about the physical grid their code is actually running on.

I realized that a massive data pipeline or machine learning training run could emit just 2 grams of CO2 per kilowatt-hour in Sweden's nuclear-heavy grid, or 708 grams in Mumbai's coal-dependent one. That’s a 350x difference for the exact same compute. Developers aren't ignoring this because they don't care; they're ignoring it because the tooling is entirely blind to it. The infrastructure to fix this already exists. The awareness didn't. I wanted to build a tool that finally hands developers a "carbon receipt" for their infrastructure.

What it does

GridDeploy is a zero-friction GreenOps tool that acts as a carbon router for your cloud infrastructure. It evaluates your deployment configurations and cross-references your target cloud region with carbon intensity data from global grid operators.

Instead of just showing you a dashboard that makes you feel guilty about your emissions, GridDeploy is actionable. It actively scans for the greenest available alternative region and offers a drop-in reroute. It takes an invisible, massive problem (the cloud industry's 2% global carbon footprint) and solves it with a single config change.

How we built it

I sprinted on this as a solo founder with very little sleep. The application is a full-stack JavaScript environment. The backend is powered by Node.js and Express to handle the routing logic and manage the API thresholds. For the frontend, I used Vite to ensure a fast, lightweight build.

To get this live for the hackathon, I set up a continuous deployment pipeline connecting my GitHub repository directly to Render. I also heavily leveraged AI coding assistants to help navigate the syntax and boilerplate, which allowed me to focus purely on the system architecture and the core business problem.

Challenges we ran into

Deployment was an absolute rollercoaster. At 3:00 AM, with just hours left until the deadline, the cloud deployment completely fought back.

My builds were succeeding, but the Render server kept crashing on startup. First, I had to hunt down a silent syntax error (Unexpected end of input) that broke the entire backend. The absolute second I fixed that and pushed the code, the server crashed again due to an Express routing bug (TypeError: Missing parameter name) where a hanging colon broke the API path. Debugging raw server logs in the middle of the night while watching the Devpost countdown timer was the ultimate stress test.

Accomplishments that we're proud of

I am incredibly proud that I managed to get a full-stack, deployed application live on the internet as a solo builder. Navigating cloud hosting, environment variables, and securing access tokens without leaking them was a massive hurdle that I successfully cleared.

Beyond the code, I'm proud of the narrative. GridDeploy proves that impactful Climate Tech doesn't have to be complicated—it just has to meet developers where they already are

What we learned

I learned that writing the code is only half the battle; deploying it to the cloud is a completely different beast. I learned the hard way how to secure API tokens in environment variables, how to force manual server deploys when webhooks fail, and how to read stack traces to figure out exactly why a server is refusing to wake up.

What's next for Grid-Deploy

Right now, GridDeploy is a powerful proof of concept. The immediate next step is building native integrations directly into CI/CD pipelines. I want to build a GitHub Pull Request Bot that automatically comments on infrastructure code, warning developers if they are deploying to a dirty grid and offering a 1-click commit to switch to a greener one.

I also want to introduce "Carbon ROI" metrics—proving to Enterprise CFOs that routing to high-renewable grids often lowers AWS billing costs at the same time. Save the planet, save the budget.

Built With

• express.js
• github
• javascript
• node.js
• render
• sqlite
• vite