Machine Learning Research Engineer (World Models)

Location: In-person in Zurich, London, or the Bay Area; most of the team currently works in person in Zurich

Company

hirify.global is an AI lab developing causal, multimodal world models that predict and interact with the world over long horizons for applications including robotics, science, healthcare, education, gaming, and defense.

What you will do

• Develop interactive, action-conditioned diffusion world models for robotics and model-based reinforcement learning.
• Implement state-of-the-art diffusion, robot-learning, reinforcement-learning, and preference-based fine-tuning algorithms.
• Design losses, reward functions, conditioning schemes, and experiments to improve interactivity and policy performance.
• Run high-cadence ablation studies across architectures, features, data mixes, and policies.
• Build data pipelines, fine-tune world models, and deploy policies on Robots-as-a-Service and partner robot systems.
• Optimize machine-learning training and inference on modern GPUs and maintain core ML frameworks.

Requirements

• PhD or equivalent research experience plus 2+ years of relevant research or engineering experience, or 4+ years of software engineering experience with 2+ years of relevant machine-learning work.
• Hands-on experience with diffusion models, world models, model-based reinforcement learning, or vision-language-action policies.
• Experience owning projects end to end and working across the full ML pipeline, from data to real-robot deployment.
• Proficiency with PyTorch, TensorFlow, or JAX.
• Highly experiment-driven approach to research and engineering.
• Flexibility to work in person in Zurich, London, or the Bay Area.

Nice to have

• Sim-to-real experience.

Culture & Benefits

• Work at the frontier of interactive world models and physical AI.
• Collaborate with researchers and engineers experienced in language models, video models, world models, and autonomous systems.
• Use real robots and partner systems as part of the research and deployment process.
• Work in a highly experimental environment where failed experiments are used to guide further research.