World Model Research Scientist for Robotics

Firstintelligence; San Francisco (CA); On-site; USD 140,000 - 210,000; Full time

Job summary

Firstintelligence is building a foundational, physics-based world model and the surfaces that ride on it, including robotics, worlds, drama, and agents. We seek a Research Scientist to set technical direction, bridging ML research and engineering with a long-horizon bet.

You should have experience with 3D Gaussian Splatting or large-scale generative models, and a CS background or equivalent depth, by degree or by doing. Join us in crafting core models and deploying them to robotics platforms.

Qualifications

• Experience with 3D Gaussian Splatting or large-scale generative models.
• Background in 3D GS, generative world models, or robotics learning.
• CS background or equivalent depth — by degree or by doing.

Responsibilities

• Set the technical direction for the world model.
• Mix of ML research, engineering taste, and a long-horizon bet.
• Train and ship the core models; own training infra, evaluation, and the loop from paper to shipped model.
• Bring the world model inside robotics platforms: perception, prediction, policy.
• A CS background or equivalent depth — by degree or by doing.
• High agency, fast learner, kind & collaborative, ownership, composure, curious yet skeptical.

Skills

3D Gaussian Splatting

Large-scale generative models

Education

CS background or equivalent

Job description

Firstintelligence is building a foundational, physics-based world model and the surfaces that ride on it, including robotics, worlds, drama, and agents. We seek a Research Scientist to set technical direction, bridging ML research and engineering with a long-horizon bet.

You should have experience with 3D Gaussian Splatting or large-scale generative models, and a CS background or equivalent depth, by degree or by doing. Join us in crafting core models and deploying them to robotics platforms.