Research Scientist

Research Scientist – Robotics & Physical AI

📍 Zurich, Switzerland

🏢 Hybrid | Full-time

✈️ Relocation support available

I’m looking for a Research Scientist to join a robotics company developing foundation models using data from robots already operating in commercial environments.

This is an opportunity to take research through to deployment, using real robot behaviour and failures to improve how models learn, adapt and perform.

There’s scope to focus on either pretraining or post-training, depending on your strengths.

What you’ll work on:

• Develop transformer and diffusion architectures that learn from visual observations, robot actions and state data.
• Shape datasets, training objectives and experiments to understand what improves generalisation.
• Work alongside ML infrastructure engineers to scale training efficiently.
• Adapt policies through imitation learning, reinforcement learning, fine-tuning and distillation.
• Test models on physical robots, investigate failures and improve reliability in unfamiliar situations.

Who I’m looking for:

• A strong research background combining machine learning with practical robotics.
• Experience building or adapting large deep learning models, particularly VLMs, VLAs or learned robot policies.
• Strong PyTorch or JAX skills and the engineering ability to troubleshoot complex systems.
• An understanding of perception, control and state estimation, alongside confidence working with hardware.
• An experimental mindset: someone who enjoys testing ideas, learning from results and getting models to work outside the lab.

Experience deploying learned policies on real robots would be particularly valuable.

You’ll join a focused team where your research can directly influence the performance of a deployed robot fleet, with flexible working hours and an English-speaking environment.

Interested? Send me a message to find out more.

#Robotics #PhysicalAI #MachineLearning #ResearchScientist #Zurich