Remote World-Model Research Scientist

Geolava; United States; Remote; USD 170,000 - 230,000; Full time

Benefits offered by this job

Competitive salary

Medical, dental, vision insurance

Retirement plan

Disability & life insurance

Job summary

Geolava is building spatial intelligence systems and seeking a PhD-level researcher to advance world-models with multi-temporal imagery and signals. You will train, evaluate, and refine predictive models, and work with AI and engineering teams to deploy capabilities onto Geolava's platform.

The role emphasizes rigorous experimentation, publication-ready documentation, and staying at the cutting edge of self-supervised and generative modeling for remote sensing data.

Qualifications

• World-model research experience with learned dynamics or related work in academia or industry.
• PhD in machine learning, computer vision, or related quantitative field.
• Fluent in deep learning for large vision models and foundation models.
• Strong experimental design with proper baselines and splits.
• Ability to operate in ambiguous, small-team settings and drive work to completion.
• Clear written and spoken technical communication for research and non-research audiences.

Responsibilities

• Design, train, and evaluate world models predicting asset changes over time and under interventions.
• Develop representation learning and predictive modeling on multi-temporal aerial/satellite data and signals.
• Build rigorous evaluation with baselines, temporal splits, calibration, and leakage-proof tests.
• Run experiments on cloud GPUs with emphasis on throughput and cost, turning results into decision-ready writeups.
• Collaborate with AI and engineering teams to deploy research into Geolava's platform.
• Keep current with world-models, self-supervised learning, and generative modeling literature.

Skills

World-model research

Deep learning

PyTorch

Vision transformers

Diffusion models

Python

Experiment design

Scientific communication

Education

PhD in machine learning, CV, or related field

Tools

PyTorch

OpenCV

GIS tooling

Cloud GPUs

Job description

Geolava is building spatial intelligence systems and seeking a PhD-level researcher to advance world-models with multi-temporal imagery and signals. You will train, evaluate, and refine predictive models, and work with AI and engineering teams to deploy capabilities onto Geolava's platform.

The role emphasizes rigorous experimentation, publication-ready documentation, and staying at the cutting edge of self-supervised and generative modeling for remote sensing data.