AI Research Scientist at London-based applied AI robotics lab founded by Magic Pony co-founder

Join an elite engineering team from Amazon AGI and InstaDeep building general-purpose robotics for the real world. Rather than iterating in simulation, you will develop 3D perception and robot learning models (VLA/RL) grounded in data from actual supermarket deployments, bridging the gap between cutting-edge AI research and physical impact.

• Develop and deploy cutting-edge 3D perception models to help robots understand and navigate chaotic, dynamic human environments.
• Research and implement robot learning approaches, including Vision-Language-Action (VLA) models and Reinforcement Learning for complex manipulation.
• Publish original research at top-tier venues like ICRA, CoRL, and NeurIPS while ensuring models are robust enough for real hardware.

• Holds a PhD or equivalent research depth with 5+ years of experience in AI robotics, computer vision, or reinforcement learning.
• Demonstrates a proven track record of deploying sophisticated models onto physical hardware, overcoming sensor noise and real-world edge cases.
• Possesses strong software engineering skills, with the ability to prototype rapidly in Python and ship production-grade code without constant oversight.

• Work directly under the founder of Magic Pony (acquired by Twitter for ~$150M) and a world-class team from Amazon, Imperial, and Cambridge.
• Access unique, high-fidelity data from robots already deployed in messy, real-world human environments rather than just lab-based simulations.
• Benefit from significant early-stage equity and competitive compensation while shaping the UK's leading applied physical AI startup from the ground floor.