Scale physical intelligence beyond the simulation!
about the company
We represent a hyper-growth deep-tech startup revolutionizing industrial automation through Embodied AI. Build generalizable VLA models and spatial intelligence platforms that control complex physical machinery in the real world
about the role
Bridge the gap between frontier AI research and physical execution by training and deploying cutting-edge VLA architectures directly onto real hardware.
- VLA & Policy Design: Design, fine-tune, and deploy Vision-Language-Action (VLA) models, behavior cloning, and RL policies for robotic manipulation and mobility.
- Sim-to-Real Transfer: Solve core sim-to-real gap challenges using real-world telemetry, multimodal sensor fusion, and rapid hardware iteration.
- Scale Infrastructure: Orchestrate multi-modal data pipelines and multi-node GPU training for large-scale foundation model experiments.
skills and experience
- MSc/PhD in Robotics, AI, or Computer Science with strong research or industry experience (1-2 years)
- Deep fluency in PyTorch, VLA architectures, vision-conditioned neural control, and policy learning and agent-based training paradigms.
- Proven track record deploying simulation-trained algorithms onto functional robotic hardware.
- Mandarin speaking required as you need to liaise with Chinese counterparts who can only speak and write in Mandarin
- Bonus: Multimodal foundation models (vision/language/kinematics) or top-tier publications (ICRA, IROS, NeurIPS, CVPR)
To apply online please use the apply function, alternatively you may contact Evangeline. (EA: 94C3609/ R24124002)
