Machine Learning Performance Engineer

The role involves optimising the performance of machine learning models through a comprehensive understanding of systems programming, GPU architecture, and high-performance computing in a fast-paced trading environment. Responsibilities include troubleshooting training and inference performance issues, enhancing CUDA and system-level processes, and integrating advanced networking technologies for GPU clusters. Candidates should possess knowledge of modern ML techniques, CUDA tools, and high-speed networking such as Infiniband and NVLink, as well as distributed training frameworks like NCCL or MPI. The position requires inventive problem-solving, analytical thinking, and fluency in English. Technical expertise in GPU programming, system optimisation, and network configurations is essential for this role.