We are seeking a Senior Deep Learning Engineer to build the future of Embodied AI. This is a role for those who work "in the weeds" of model architecture and training loops not API wrappers or prompt engineers.
The Mission
Architect Behavior: Own the development of motor policies using Behavior Cloning and RL.
Scale VLA Research: Lead pre/post-training on our Vision-Language-Action (VLA) stack.
Engineered Data: Build automated pipelines to ingest teleop logs/synthetic data, apply weak-supervision, and curate high-quality datasets.
Failure Analysis: Systematically identify and solve failure modes through retraining loops.
Technical Requirements
Experience: 3+ years in deep learning (shipped models, research papers, or major OSS).
Domain Depth: Hands-on expertise in VLMs, Diffusion/Generative Video, or LLM Pre-training.
Frameworks: Mastery of Python and PyTorch or JAX.
Systems: Experience with distributed training and complex numerical debugging.
Randstad Technologies Ltd is a leading specialist recruitment business for the IT & Engineering industries. Please note that due to a high level of applications, we can only respond to applicants whose skills & qualifications are suitable for this position. No terminology in this advert is intended to discriminate against any of the protected characteristics that fall under the Equality Act 2010. For the purposes of the Conduct Regulations 2003, when advertising permanent vacancies we are acting as an Employment Agency, and when advertising temporary/contract vacancies we are acting as an Employment Business