Internship - Humanoid Motion Generation (Diffusion or Flow Matching)

Location: Zürich, Switzerland

Company

hirify.global is building the intelligence layer for next-generation humanoid robots, focusing on reinforcement learning and real-world humanoid deployment.

What you will do

• Develop generative whole-body trajectory generation for humanoid robots.
• Apply diffusion models, flow matching, and autoregressive models to robot motion generation.
• Work on 3D environment perception and the interaction of humanoid robots with the world.
• Implement state-of-the-art methods from recent research literature.
• Contribute to foundational robotics software through rapid experimentation and clean, modular code.

Requirements

• Currently pursuing a Master’s or PhD in Computer Science, Robotics, Machine Learning, or a related technical field.
• Project experience with diffusion models, flow matching, or learning-based trajectory generation.
• Familiarity with generative modeling applied to time-series data.
• Excellent Python and PyTorch skills.
• Strong analytical thinking and problem-solving skills.
• Ability to implement state-of-the-art methods from recent research literature.

Nice to have

• Research or project work in human motion generation, neural avatars, robotic imitation learning, or egocentric action prediction.
• Experience with robotics or simulation environments.
• Experience with multimodal generative models.
• Experience fine-tuning foundation models such as Gr00t or SmolVLA for whole-body actions or kinematic motions.

Culture & Benefits

• Competitive compensation.
• Opportunity to contribute to an ambitious European robotics company.
• Energetic, collaborative environment with a bias for action.