Machine Learning Performance Engineer

A machine learning performance engineer specialises in optimising the training and inference of models within a high-frequency trading environment, utilising low-level systems programming and hardware understanding. The role involves enhancing large-scale training efficiency, low-latency real-time inference, and high-throughput research inference, with emphasis on CUDA, storage systems, and network integrations. Candidates should possess skills in debugging and system optimisation using specialised tools, technical knowledge of GPU architecture, and experience with networking technologies linking GPU clusters. The position requires familiarity with distributed training algorithms and a proactive approach to problem-solving. Fluency in English is essential, and the role offers opportunities to innovate across hardware, software, and networking layers.