Overview

hum.ai develops AI software to measure ocean health by fusing satellite imagery and ground-truth data with multimodal geospatial models. It trains on the SeeFar dataset to combine historical data with high-resolution imagery, enabling detailed analytics for conservation, carbon removal, and government monitoring. The models work across multiple satellite types and resolutions, using a large curated dataset to avoid being tied to a single satellite. The long-term goal is a nature-focused AGI that understands and helps manage the natural world, starting with ocean ecosystems to support conservation and environmental policy.