AI Deforestation Early Warning Developer

Kelvix Systems

Remote · $130k – $160k / year · Full-time · Posted 5 months ago

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

About the role

By the time a deforestation event appears in a quarterly satellite mosaic, thousands of hectares may already be gone. Near-real-time detection from Sentinel-2 and Planet imagery — revisit cycles of 2 to 5 days — can catch clearing events within days of their start, enabling early intervention by conservation organizations and government enforcement agencies. We build the satellite-based forest monitoring system used by 12 conservation organizations and 4 national forestry agencies. Our change detection pipeline analyzes multispectral imagery across 80 million hectares of tropical forest continuously. The model achieves 89% recall on clearing events larger than 5 hectares; the priority is improving sensitivity to incremental clearing. You will redesign the change detection approach for small clearing events: better handling of cloud cover in tropical regions, temporal fusion across multiple passes to detect gradual clearing, and uncertainty quantification that lets the operations team prioritize their limited verification capacity.

Fully remote.

Responsibilities

Improve change detection sensitivity for 0.5 to 5 hectare clearing events
Build cloud masking and gap-filling approaches for tropical multispectral imagery
Implement temporal fusion across multiple satellite passes for gradual clearing detection
Design uncertainty quantification for change detection results
Create evaluation benchmarks validated against field verification ground truth
Collaborate with conservation organizations on operational deployment requirements

Requirements

5+ years in satellite remote sensing or geospatial ML with change detection experience
Experience with multispectral satellite imagery: Sentinel-2, Landsat, or Planet
Strong PyTorch skills and familiarity with geospatial deep learning libraries
Understanding of tropical forest ecology and clearing patterns
Ability to work with mission-driven conservation stakeholders