Inspiration

We kept seeing news about floods, heatwaves, and extreme weather, but most of the information around climate risk felt either too technical or arrived too late to be useful. Climate data exists everywhere, yet it’s rarely presented in a way that helps people take action. That gap inspired us to build Indradhanu a platform that makes climate risk information easy to understand, timely, and practical, so people can prepare instead of reacting after damage is done.

What it does

Indradhanu helps users understand climate risks in real time and what they can do about them. It shows current conditions, predicts short-term risks, and alerts users when something serious is developing. Users can ask questions using voice, track risks in different locations, and explore how different adaptation steps can reduce climate impact. The focus is not just on data, but on clear signals and useful guidance.

How we built it

We built Indradhanu as a web-based platform using modern frontend tools and lightweight AI models. Real-time climate data is collected from public weather APIs and processed to generate risk scores. Machine learning models help predict near-term risks, while interactive maps and charts make patterns easy to see. We also added voice interaction and alert systems to make the platform more accessible and practical for everyday use.

Challenges we ran into

Working with real-time climate data was challenging because it’s noisy and constantly changing. Designing risk scores that actually make sense to users took several iterations. We also faced challenges with performance while running AI models in the browser and making voice features work smoothly across devices. Each challenge pushed us to simplify and improve the system.

Accomplishments that we're proud of

Building a complete climate intelligence platform from scratch

Turning complex climate data into clear, usable insights

Making the platform accessible through voice interaction

Creating something that feels useful beyond just a demo

Working together effectively under tight timelines

What we learned

This project taught us how difficult — and important — it is to make technical systems understandable to non-technical users. We learned how to work with climate data, deploy AI responsibly, and design features that prioritize real-world usefulness over complexity. Most importantly, we learned how to collaborate and adapt quickly when things didn’t work the first time.

What's next for Indradhanu

Next, we want to improve prediction accuracy, add more locations, and support longer-term climate trends. We also plan to expand alert delivery options and build tools specifically for communities, planners, and emergency teams. Our goal is to grow Indradhanu into a platform that people can rely on when climate risks matter most.

Built With

  • api
  • browser
  • css
  • framer
  • git
  • gps
  • motion
  • next.js
  • notifications
  • open-meteo
  • plotly.js
  • shadcn/ui
  • speech
  • tailwind
  • tensorflow.js
  • typescript
  • vercel
  • web
Share this project:

Updates