Research Scientist, Biosphere Models

Design, implement, and train state-of-the-art Geospatial AI models on planetary-scale datasets, with emphasis on forests and habitats.
Develop self-supervised or weakly-supervised pretraining methods to address data scarcity in natural environments.
Build scalable data pipelines for ingesting and processing heterogeneous Earth Observation data.
Lead technical validation against ground truth, designing validation campaigns and geospatial annotation strategies.
Collaborate with domain experts to align model objectives with downstream ecological and conservation applications.
Communicate findings clearly, publish results, and contribute to open-source code releases and team goals.