Project Overview
Wildfire Insight Intelligence is an AI-powered platform for assessing and classifying wildfire damage using satellite imagery. Built for Brazil Flying Labs, the system analyzes Sentinel-2 satellite data through Google Earth Engine to evaluate burn severity using multiple metrics (dNBR, RBR, dNDVI). AI-generated reports powered by Google Gemini provide detailed analysis, and users can ask follow-up questions through an interactive conversation interface. The platform runs on 6 platforms with a consistent experience across web, mobile, and desktop.
The Challenge
Assessing wildfire damage across ecological reserves required manual field surveys that were slow, dangerous, and limited in scope. Researchers needed a way to rapidly evaluate burn severity across large areas using satellite data, generate standardized reports, and collaborate across teams — accessible from any device whether in the office or in the field.
Our Solution
We built a full-stack platform with a Django REST API that integrates Google Earth Engine for satellite imagery analysis and Celery for asynchronous processing of assessment jobs. The backend stores geospatial data in PostgreSQL with PostGIS and uses Google Gemini AI to generate detailed analysis reports with interactive Q&A. Three separate frontends — React web app, React Native mobile app (iOS/Android via Expo), and Electron desktop app — provide a consistent experience across all 6 platforms. Real-time push notifications keep users updated on assessment progress.