Hudson River Trading (HRT) is seeking a lead Software Engineer with a specialization in performance engineering to optimize our high-performance and AI compute workloads. This role will apply broad engineering skills across infrastructure, tools, and applications to optimize the performance of our most critical workloads. This will allow us to accelerate innovation and update our strategies faster, a crucial component of HRT's continued success in advanced automated trading.
Successful candidates will operate across multiple stacks and synthesize a holistic view of challenges and potential solutions. This role is well-positioned to be successful within HRT's collegial and non-siloed environment as it involves working closely with a wide range of stakeholders and teams to make our high-perf applications as efficient as possible.
Responsibilities
- Investigate and measure the performance of AI and other high-performance compute workloads
- Optimize workload performance by designing and implementing solutions across infrastructure, tools, and application code
- Track the performance of workloads over time and identify potential areas for long-term investments and improvements
- Partner with users and platform engineers to understand workload priority and ensure all stakeholders are aligned with suggested plans
- Identify, hire, and mentor engineers to accelerate the performance improvement work
Qualifications
- Expert level C++ programming ability, including low-level
- 10+ years of relevant experience in HPC/AI Performance Optimization
- Demonstrated ability to optimize complex workloads across multiple software stacks
- Experience with AI frameworks (PyTorch, TF, MXNet, etc.) and accelerated programming stacks (CUDA, MPI, AVX, or similar)
- Quantitative/data-oriented mindset
- Great communication skills, with the ability to effectively communicate technical ideas to a variety of stakeholders across the firm
- Track record of leading complex projects end-to-end
- Proven ability to lead, mentor, and strategically grow successful teams