Staff Engineer - World Models & Multimodal AI

Odyssey; Greater London; On-site; GBP 70,000 - 110,000; Full time

Job summary

Odyssey in the United Kingdom is seeking deeply technical researchers to push world models, video generation, and multimodal ML toward real applications that matter.

You will prototype ideas from hypothesis through experiment to working code, emphasizing conditioning, controllability, and long-horizon consistency.

Join a world-class team with ties to academia and industry and help define how AI perceives and interacts with pixels over the next decade.

Qualifications

• We seek deeply technical staff with a track record in world models or related areas.
• Experience shipping systems or research at scale is valued.
• Ability to define problems, design experiments, and build minimal viable systems.

Responsibilities

• Push world models beyond current diffusion and transformer limits with new architectures or training paradigms.
• Train conditioning methods and control policies for real-world generation.
• Take ideas from hypothesis to prototype grounded in concrete applications.
• Design experiments, distributed training, and inference pipelines to enable breakthroughs.
• Define measurable criteria for controllability and eval performance.

Skills

World models

Generative modeling

Video generation

Multimodal learning

Large-scale ML systems

Conditioning

Controllability

Experiment design

Cross-functional collaboration

Tools

Python

TensorFlow

PyTorch

Job description

Odyssey in the United Kingdom is seeking deeply technical researchers to push world models, video generation, and multimodal ML toward real applications that matter.

You will prototype ideas from hypothesis through experiment to working code, emphasizing conditioning, controllability, and long-horizon consistency.

Join a world-class team with ties to academia and industry and help define how AI perceives and interacts with pixels over the next decade.