AI Research Engineer, Pre-Training

Sourced from jobs-radar.com

Hudson River Trading is hiring an AI Research Engineer for pre-training models used in trading. The role involves improving large-scale deep learning training systems, including kernel development and distributed training. You will collaborate with researchers to co-design and optimize models for market prediction. Requires strong engineering skills in CUDA, PyTorch, and related technologies.

• Improve all aspects of large-scale model training including kernel development, training data loading, parallelism, networking, and fault-tolerance. Work closely with researchers to co-design and improve models, shaping the research agenda.

• Strong engineering skills in CUDA, Triton, Pallas, CuTe, PyTorch, JAX, XLA, CUDA Graphs, FPGA/ASIC. Two or more years of experience building deep learning systems in any domain.

• LLM experience, experience translating methods between application areas.