AI Research Engineer, Inference
Hudson River Trading5 months ago
Base Salary
$250k - $300k/yr
Responsibilities
- Improve all aspects of large-scale model inference, including GPU kernels, novel inference devices, and data streaming.
- Develop solutions using off-the-shelf and in-house ASICs and FPGAs.
- Collaborate with researchers to co-design and improve models and help shape the research agenda.
- Optimize inference systems and devices deployed globally for varied trading needs and model types.
Requirements
- Strong engineering skills, particularly in CUDA, Triton, Pallas, CuTe DSL kernel development, lower-level PyTorch, JAX, XLA, CUDA Graphs, or FPGA/ASIC development.
- At least two years of work experience building deep learning systems in any application domain.
- Experience translating methods between application areas is highly valued.
- LLM experience is valuable but not required.
- Finance experience is not required.
Benefits
- Competitive benefits package
- Eligible for discretionary performance-based bonuses
Tech Stack
Categories
AI ResearchML Engineering
About Hudson River Trading
Hudson River Trading is a privately held quantitative trading firm that researches and builds automated trading algorithms and low-latency systems. It trades its own capital across global electronic markets in multiple asset classes, often acting as a liquidity provider. Founded in 2002 and headquartered in New York, the firm runs a large-scale trading platform with around-the-clock operations supporting live market activity and risk controls.