ML Systems Scientist — High-Throughput Inference & GPUs

Jobtailor; California (MO); On-site; USD 150,000 - 210,000; Full time

Job summary

Jobtailor is seeking a senior AI systems engineer to design and optimize high-throughput, low-latency inference systems, write GPU kernels, and mentor teams. You will benchmark with PyTorch Profiler, collaborate with infrastructure to build scalable distributed systems, and translate measurements into performance improvements.

The role requires a Master’s/PhD in CS or EE and deep knowledge of GPU architecture, Python, C++, CUDA, and Triton.

Qualifications

• Master’s or PhD in CS, EE, or related field.
• Hands-on experience scaling large-scale inference or serving workloads.
• Strong understanding of GPU architecture.
• Proficiency in Python and C++.
• Familiarity with CUDA or Triton for performance‑critical workloads.
• Demonstrated ability to make engineering decisions based on rigorous measurement and benchmarking.

Responsibilities

• Design and optimize high‑throughput, low‑latency inference systems
• Write and maintain high‑performance GPU kernels
• Conduct deep performance analysis using tools such as PyTorch Profiler
• Partner with infrastructure teams to design scalable distributed systems
• Establish and track efficiency metrics and build benchmarking frameworks
• Serve as a trusted technical advisor to research and product teams

Skills

GPU kernels

Python

C++

GPU architecture

high-throughput systems

low-latency systems

benchmarking

measurement

Triton

CUDA

PyTorch

Education

Master's in CS

PhD in CS

Master's in EE

PhD in EE

Tools

CUDA

Triton

PyTorch

Job description

Jobtailor is seeking a senior AI systems engineer to design and optimize high-throughput, low-latency inference systems, write GPU kernels, and mentor teams. You will benchmark with PyTorch Profiler, collaborate with infrastructure to build scalable distributed systems, and translate measurements into performance improvements.

The role requires a Master’s/PhD in CS or EE and deep knowledge of GPU architecture, Python, C++, CUDA, and Triton.