Similar Jobs

See all

About Kog:

  • We build real-time LLM inference for AI agents with a focus on low latency.
  • Our engineers co-design model architecture and GPU kernels using low-level hardware optimizations.

What You'll Work On:

  • Develop and optimize GPU kernels using CUDA, HIP, and low-level instructions.
  • Profile and benchmark GPU workloads to identify bottlenecks and validate improvements.
  • Help build AI agents that automate GPU optimization and correctness testing.

What We're Looking For:

  • Strong experience in GPU kernel development using CUDA or HIP/ROCm.
  • Deep understanding of GPU architecture, memory hierarchy, and parallel execution.
  • Proven ownership of complex technical projects from implementation to results.

Location and Availability:

  • Must already be based in Paris or Île-de-France.
  • We are not considering relocation from other regions.
  • Looking for someone who can join the team quickly.

Kog

At Kog, we're building real-time LLM inference for AI agents. Our engineers work directly with GPU hardware, developing optimized kernels and AI agents to accelerate GPU engineering.

Apply for This Position