EcoInsight

An intelligent platform that transforms complex environmental datasets into interactive, personalized climate stories and actionable insights for communities worldwide.

Comment

Inspiration

Climate change is one of the most pressing challenges of our time, yet data about it can often feel abstract or overwhelming. We wanted to create a platform that could turn raw climate data into engaging and personalized climate narratives—something that speaks not just to the mind, but to the heart. EcoInsight was born out of the desire to empower individuals and communities with meaningful stories that inspire awareness and action.

What it does

EcoInsight is an AI-powered web platform that transforms real-time environmental data—like weather, air quality, and climate indicators—into dynamic, human-readable climate stories. Users can enter any city in the world and instantly receive a personalized story about that location's environmental condition, crafted using cutting-edge generative AI and live data sources like OpenWeatherMap.

How I built it

  • Frontend: Built with Next.js and Tailwind CSS for a modern, responsive, and accessible UI.
  • Backend: Developed with FastAPI, using asynchronous APIs to fetch real-time weather and air quality data.
  • Database: MongoDB is used to store generated climate stories along with embeddings for semantic search.
  • AI Integration: Leveraged Google Vertex AI’s Gemini 1.5 model to generate insightful and readable climate stories based on structured data.
  • Embedding & Search: SentenceTransformers are used to embed stories for similarity-based retrieval.
  • Deployment: Hosted backedn on Google Cloud Platform using Cloud Run, and frontend using Vercel.

Challenges I ran into

  • Integrating multiple APIs (weather, air quality, and Vertex AI) and ensuring real-time performance.
  • Ensuring the AI-generated content was not only factually aligned with the data but also readable and emotionally resonant.
  • Designing a scalable backend that could handle asynchronous tasks while remaining responsive to frontend interactions.
  • Creating embeddings and vector storage in MongoDB for efficient semantic search.

Accomplishments that I am proud of

  • Successfully generated unique climate stories using real-time data and generative AI.
  • Enabled global location support through a flexible search interface.
  • Created an aesthetically pleasing and intuitive UI that makes data approachable.
  • Built a scalable backend that seamlessly integrates AI, APIs, and a vector database.

What I learned

  • The power of combining generative AI with real-world data to drive awareness.
  • How to integrate Google Vertex AI into a full-stack application.
  • Effective use of MongoDB with vector embeddings for semantic search.
  • How to optimize user experience through storytelling rather than raw metrics.

What's next for EcoInsight

  • 🌍 Add satellite-based climate anomaly tracking (e.g., deforestation, droughts).
  • 💬 Add multilingual support for stories to reach more communities.
  • 📈 Incorporate trends and historical comparisons to enrich stories.
  • 🔍 Add interactive visualizations alongside each story.
  • 🧠 Let users compare their location with other regions using AI insights.
  • 🧾 Launch a sustainability dashboard for organizations and local governments.

Built With

Share this project:

Updates