Research Scientist (Embodied AI)

Location: On-site in London or Munich

Company

hirify.global develops generative AI and computer vision models that understand the physics and geometry of the physical world for robotics, AR/VR, gaming, and cinema.

What you will do

• Own the end-to-end training pipeline for vision-language-action and world-action models, from data preparation to policies running on robots.
• Set technical direction for embodied AI research.
• Close the sim-to-real gap using domain randomization, system identification, and calibration.
• Adapt vision-language model backbones for robotic control through encoder, adapter, and co-training strategies.
• Curate heterogeneous robot datasets and design action representations and decoding methods, including tokenization, chunking, diffusion, and flow matching.
• Build action-conditioned world-model components and apply supervised fine-tuning and reinforcement learning for target embodiments and robustness.

Requirements

• PhD in robotics, machine learning, or computer vision with a focus on robot learning, with relevant industry experience accepted in addition to the PhD.
• Deep experience training modern robot policies such as VLA, WAM, and diffusion models end to end.
• Strong foundations in imitation learning and familiarity with reinforcement learning fine-tuning of pretrained policies.
• Fluency with vision-language model backbones and their adaptation for control.
• Expert Python and PyTorch skills, including multi-node distributed training with FSDP or an equivalent framework.
• Publications at leading robotics, computer vision, or machine learning venues, open-source work, and/or deployed systems.

Culture & Benefits

• Work in an engineering research environment focused on generative AI, computer vision, world models, and physically grounded 3D environments.
• Inclusive workplace committed to diversity and equal opportunity.