Principal Research Engineer (World Models)

Location: New York City or San Francisco

Working Arrangement: Full-time, On-site

About the Opportunity

We’re partnering with an ambitious, venture-backed robotics AI company developing advanced autonomous systems for real-world industrial applications.

The company is focused on bringing intelligent robotic systems into complex, dynamic environments, tackling physically demanding tasks that require high levels of reliability, adaptability and precision.

With a multidisciplinary team spanning robotics, machine learning, hardware and software engineering, the business is developing its technology in-house, from physical systems through to the intelligence that powers them.

Following significant technical and commercial progress, the company is now looking to expand its research capabilities with a senior technical hire who can help shape the next generation of its AI architecture.

The Role

We're looking for a Principal AI Research Engineer specialising in World Models and Generative Modelling to drive research and development at the intersection of generative AI, simulation and robotics.

You'll play a key role in developing models that allow robotic systems to understand, predict and interact with complex physical environments.

This is a highly influential position for someone who has worked on large-scale generative models and wants to translate cutting-edge research into practical, real-world robotic capabilities.

What You'll Be Working On

• World Model Development: Architect and train large-scale generative models capable of learning complex environments and predicting future states.
• Action-Conditioned Simulation: Develop controllable generative models that can simulate how environments evolve in response to different actions and inputs.
• Model Training & Scaling: Lead research into training methodologies, model architectures and scaling approaches for next-generation world models.
• Data Infrastructure: Develop the data pipelines and infrastructure required to train, curate and improve large-scale world models.
• Evaluation Frameworks: Establish robust evaluation methodologies to measure model performance, predictive accuracy and downstream impact on robotic systems.
• Research to Deployment: Work closely with robotics and engineering teams to translate research breakthroughs into practical capabilities.

What We're Looking For

We're particularly interested in candidates who have:

• Experience architecting and training large-scale video generation models, world models or other advanced generative systems.
• Strong understanding of generative modelling, sequence modelling and modern deep learning architectures.
• Experience developing action-conditioned or state-conditioned models capable of generating controllable future predictions.
• Experience building large-scale training datasets, data pipelines and evaluation frameworks.
• A track record of taking complex research concepts through to working technical implementations.
• The ability to operate at a senior technical level, setting research direction and influencing engineering decisions.

Particularly interesting backgrounds include:

• Large-scale video generation and predictive modelling.
• Generative simulation and environment modelling.
• World models for embodied intelligence.
• Research involving controllable video generation, temporal modelling or interactive environments.

Why Consider This Opportunity?

• Technical Ownership: Take a leading role in defining the company's world-model research direction.
• Real-World Impact: Work on technology being developed for deployment in physical environments, rather than purely research or demonstration purposes.
• Strong Foundations: Join a business with established engineering capabilities and meaningful commercial traction.
• Influence: Help shape the technical roadmap and future research capabilities of a growing AI organisation.
• Equity & Upside: Opportunity to join at a stage where senior technical hires can have significant influence over the company's future.

Interested?

If you're working on world models, generative video, predictive modelling or controllable simulation and want to apply that experience to real-world robotics, please apply.

We'd particularly welcome examples of research projects, technical publications, open-source contributions or systems you've built.