Senior Deep Learning Performance Engineer

A leading quantitative trading firm is seeking a Machine Learning Performance Engineer to join a small, highly technical team running a deep learning trading strategy in live markets. The strategy is early in its development, so this is a chance to build core performance infrastructure alongside the lead researcher instead of inheriting a finished platform.You'll work at the intersection of deep learning, GPU programming, and systems engineering. Speed and efficiency of both training and inference translate directly into trading performance.What You’ll Work OnWrite and optimize custom CUDA kernels for latency- and throughput-critical deep learning workloadsBuild scalable, robust training and inference pipelinesIdentify and eliminate performance bottlenecks across the GPU, memory hierarchy, and software stackDig into the internals of open-source deep learning frameworks (PyTorch, JAX) and extend them where neededWork directly with researchers and traders to turn model ideas into fast production systemsDevelop a deep understanding of the trading systems your work supportsQualificationsHands-on CUDA kernel development, with strong knowledge of Tensor Cores, cooperative groups, CUDA Graphs, and warp-level intrinsicsStrong C++ and PythonDeep understanding of computer architecture, including memory hierarchy and GPU executionExpertise in the internals of deep learning frameworks such as PyTorch, JAX, or TensorFlowExperience building or optimizing high-performance deep learning systems in any domain (robotics, recommendations, audio, video, biology, and so on)Nice to HaveExperience with Triton, CUB, cuDNN, or cuBLASJAX ecosystem experience (XLA, Flax)Linux systems programmingLarge-scale distributed trainingOpen-source contributions to ML or GPU projectsPrior finance or trading experience is not required.