Senior Kernel Engineer (AI)

TL;DR

Senior Kernel Engineer (AI): Building and optimizing high-performance GPU kernels for next-generation AI systems with an accent on CUDA kernel design and GPU-to-GPU communication paths. Focus on profiling, debugging memory bottlenecks, and maximizing performance-per-watt for large-scale model inference.

Location: Santa Clara, CA; New York City, NY; San Francisco, CA

Salary: $225,000–$300,000

Company

An early-stage AI infrastructure startup backed by a Tier-1 VC, innovating at the chip and system level to optimize AI inference performance.

What you will do

• Design, implement, and optimize CUDA kernels for performance and scalability.
• Build and tune GPU-to-GPU communication paths, including NCCL-style collectives and P2P.
• Profile, debug, and optimize bottlenecks related to memory, latency, and throughput.
• Collaborate closely with compiler, systems, and hardware teams to optimize the AI stack.

Requirements

• 3+ years of experience in kernel development and performance optimization.
• Deep understanding of GPU architecture, memory hierarchies, and execution models.
• Experience with multi-GPU communication and synchronization.
• Must be located in or able to work from Santa Clara, New York City, or San Francisco.

Nice to have

• Experience with Triton.
• Familiarity with AMD GPUs and ROCm.

Culture & Benefits

• Opportunity for massive ownership and high impact in an early-stage environment.
• Chance to build cutting-edge AI infrastructure from the ground up.
• Backed by Tier-1 Venture Capital.