2 of 2 Permanent CUDA Jobs in Manchester

scale ML systems, optimise distributed training and inference, and build GPU‐accelerated infrastructure in close collaboration with the founders. The role demands deep Python, CUDA C/C++, and PyTorch expertise, plus strong system design skills and ownership. #J-18808-Ljbffr ...

inference infrastructure Own model quality, performance, and reliability ML Infrastructure & Optimisation; Optimise distributed training and inference pipelines Design GPU-accelerated systems, including custom CUDA kernels when needed Profile and optimise data pipelines, training loops, inference, and deployment Build internal tooling, benchmarking systems, and evaluation frameworks Architecture & Collaboration; Design scalable … excellence across the ML platform What We’re Looking For Must Have; Extensive experience designing and building large-scale foundation models Strong Python and CUDA C/C++ expertise Deep knowledge of PyTorch (preferred) or another major deep learning framework Experience optimising and debugging deep learning models Experience with ...