Machine Learning Engineer – Physical AI / Robotics

Machine Learning Engineer – Physical AI / Robotics What if the models you trained didn’t just generate an answer, but made a robot move? This is an opportunity to work at the intersection of machine learning, robotics and Physical AI , building models that learn how to interact with the real world. You’ll join an ambitious UK AI startup at an early enough stage to have genuine influence over the technology, working on everything from large-scale model training to evaluation, data and deployment on real robotic hardware. What’s in it for you? * Founding-stage engineering role with meaningful equity * Train vision-language-action models and world models * See your models tested on real robots , not just benchmarks * Work across the full ML loop: data → training → evaluation → deployment * Build infrastructure that makes every training iteration faster and more effective * Significant technical ownership without layers of process or bureaucracy * Help shape the ML foundations of a company tackling one of AI’s hardest problems Your work will include: * Training and fine-tuning large-scale machine learning models * Building high-performance training and evaluation pipelines * Developing the data infrastructure needed for rapid experimentation * Improving dataset quality and creating better feedback loops * Optimising model performance, training efficiency and iteration speed * Evaluating models on real robotic systems * Using results from hardware testing to inform the next training cycle The core experience we’re looking for is: * Strong experience training deep learning models * Excellent Python skills and experience with modern ML frameworks * A strong understanding of model training, optimisation and performance * Experience building reliable ML training or evaluation infrastructure * Solid mathematical foundations and a rigorous approach to experimentation * The ability to diagnose why a model isn’t working and systematically improve it * An appetite for the ownership and ambiguity that comes with an early-stage company You don’t need to have spent your career in robotics. What matters is that you’re a strong ML engineer who wants to work on models that perceive, reason and act in the physical world .