⚡ Zerostrike

Zerostrike is a proactive wildfire remediation solution powered by autonomous drones controlled by AI agents. We call ourselves "the Palantir of wildfires".

We trained ML models to accurately predict where "dry lightning" will strike, and built the infrastructure and hardware to autonomously deploy drones to areas of risk, releasing cloud seeding solution to trigger rainfall and prevent disasters before they happen.

The platform pulls data from tens of sources (NASA satellites, ESA geospatial datasets, a decade of fire history, moisture readings) combined into a single data layer updated every hour. Our prediction engine (XGBoost trained on 28,000 real fire ignitions from the 2020 California Lightning Complex) calculates land risk, strike probability, and consequence scores to identify the highest-risk areas.

Claude powers the autonomous decision loop. The agent has full access to the live threat map, storm trajectories, drone fleet telemetry, and satellite imagery. When a collision is detected, it autonomously selects the optimal drone, calculates an intercept route, and dispatches it in under three seconds.

I built the drone integration layer. Since DJI exposes no direct API, I built a custom Android app on DJI MSDK v5 that intercepts the phone-to-remote control link, injecting programmatic waypoint missions. An ESP32-controlled servo mounted on a DJI Mini 4 Pro releases cloud seeding payload at calculated release points, with Firestore as the real-time message bus bridging everything across separate networks.

Built alongside Eniola, Emmanuel, and Julian at HackEurope.