Ethron

🌍 Inspiration

We were inspired by the Noosphere—Earth’s collective consciousness—and the urgent need to predict crises before they happen. While cities like Dubai use AI to forecast fires, there’s no open global system that integrates real-time data and predictive AI to anticipate environmental, health, or humanitarian disasters. Ethron was born to bridge that gap.

💡 What it does

Ethron acts as the Earth’s AI-powered nervous system. It collects real-time data from satellites, health agencies, sensors, and news APIs, then uses advanced AI models to predict emerging threats like wildfires, disease outbreaks, or climate anomalies. A 3D neural globe UI visualizes these risks while GPT agents suggest mitigation actions and simulate "what-if" scenarios for users.

🛠️ How we built it

• Frontend: React + Tailwind CSS + Three.js
  
• Backend: Python (FastAPI), deployed on Google Cloud (Cloud Run)
  
• Database: MongoDB Atlas with Vector Search
  
• AI Models: OpenAI GPT-4o, Gemini 1.5 Flash (Vertex AI)
  
• DevOps: GitLab CI/CD
  
• APIs/Data Sources: NASA Earthdata, WHO, Firewatch API, NewsAPI, UN datasets
  
• Real-time data pipelines were handled via Google Cloud Pub/Sub and contextual GPT prompts were designed for ethical AI assistance.

🚧 Challenges we ran into

• Aligning real-time, multi-source data feeds
  
• Ensuring ethical, explainable AI predictions
  
• Designing a globe UI that feels intelligent and humane
  
• Managing low-latency vector searches and semantic retrieval in MongoDB
  
• Preventing GPT hallucinations in critical decision outputs

🏆 Accomplishments that we're proud of

• Created a working prototype of a global-scale crisis prediction engine
  
• Built a vector-search-powered MongoDB knowledge base for real-time inference
  
• Designed an intuitive 3D globe UI with contextual AI agents
  
• Integrated real-world data pipelines and GCP infrastructure seamlessly

📚 What we learned

• How to build socially responsible, large-scale AI systems
  
• The importance of fast, meaningful vector search in crisis data
  
• How to balance powerful generative models with real-time integrity and trust
  
• Designing AI UX for high-stakes global use cases

🚀 What's next for Ethron

• Expand data coverage to include financial, geopolitical, and supply chain risks
  
• Enable citizen-reporting integrations to crowdsource early signals
  
• Collaborate with NGOs and governments to deploy alerts in underserved regions
  
• Add multilingual GPT agents and AR-based risk visualization
  
• Open-source a lightweight SDK for researchers to contribute localized risk models

Ethron: The AI nervous system of Earth — because the planet should think before it breaks.

Built With

• ai
• fastapi
• firebase
• firewatch
• gemini-1.5-flash
• gitlab
• google-cloud
• javascript
• mongodb
• nasa-earthdata
• newsapi
• node.js
• python
• react
• tailwind-css
• un-data-hub
• vertex
• who