VLA/WAM Research Engineer

About the role

Flexion builds the intelligence layer behind the next generation of humanoid robots, founded by scientists in robot reinforcement learning and backed by international VC firms. As VLA/WAM Research Engineer in Zurich, you will develop the foundation models that power humanoid robots, covering model architecture, large-scale pre-training, adaptation and post-training.

Key responsibilities

• Design and develop multimodal foundation models combining vision, language, robot state, actions and temporal context for general-purpose robotic control
• Run large-scale pretraining across diverse robot embodiments, teleoperation, egocentric human data and simulated experience
• Improve data mixtures, training objectives, optimization strategies and distributed training workflows
• Develop predictive world and action models for representation learning, future-state prediction, planning and policy improvement
• Create methods to adapt pretrained models to new tasks, environments and embodiments
• Use real-world and simulated robot experience, including imitation learning, RL, reward/value-guided learning and specialist-to-generalist distillation
• Build methods to learn from rollouts, failures and interventions while avoiding catastrophic forgetting
• Design experiments and evaluate models in simulation and on real robots

Required profile

• Research or engineering experience in VLA/VLM architectures, multimodal foundation models, world models, large-scale pretraining, robot learning or RL/post-training
• Experience owning substantial model development or training experiments, including architecture, objectives, optimization and evaluation
• Solid understanding of deep-learning architectures, representation learning and the link between data, objectives and performance
• Ability to design experiments, diagnose training failures and evaluate recent research
• Comfortable working across architecture, training and adaptation with particular expertise in at least one area

Required skills

• Strong Python and PyTorch skills, with hands-on experience training deep-learning models
• Plus: distributed training, multi-node GPU systems, reinforcement learning, offline RL, continual learning, and experience with humanoid robot learning and large-scale embodied datasets

What we offer

• Competitive compensation and enhanced pension plan
• Relocation and permit sponsorship
• Enhanced holiday and paid leave perks
• Central Zurich office with robotics testing facilities