Senior GPU Compute & AI Kernel Engineer

About the role

We are seeking a Senior GPU Compute & AI Kernel Engineer to work at the performance‑critical layer of the AI stack. You will design, implement, and optimise low‑level compute kernels that power modern deep‑learning workloads on GPUs and AI accelerators.

Key responsibilities

• Design, implement and optimise high‑performance compute kernels using modern C++ and one or more of CUDA, SYCL, OpenCL, Triton or Apple Metal.
• Profile, tune and debug parallel algorithms for throughput, latency and memory efficiency.
• Accelerate PyTorch and TensorFlow workloads by creating custom operators for Transformers, LLMs and CNNs.
• Map deep‑learning operators onto target accelerator architectures, leveraging tensor units and memory hierarchies.
• Collaborate with compiler engineers on LLVM/MLIR based kernel code generation.
• Partner with hardware teams to evaluate new accelerators and guide co‑design decisions.
• Maintain benchmarking, regression tracking and profiling methodology.
• Contribute to accelerator runtimes and framework integrations, following Linux development best practices.

Required profile

• 5+ years of software development experience with modern C++ (C++17/20 preferred).
• Strong background in high‑performance compute kernel development.
• Hands‑on GPU programming experience with CUDA, SYCL, OpenCL, Triton or Apple Metal.
• Solid understanding of GPU and AI accelerator architectures.
• Experience optimising parallel algorithms for performance and memory efficiency.
• Experience developing or optimising machine‑learning workloads in PyTorch or TensorFlow.

Required skills

• C++ (C++17/20)
• CUDA, SYCL, OpenCL, Triton, Apple Metal
• PyTorch, TensorFlow
• LLVM, MLIR
• Linux development environment, Git, CI/CD pipelines
• Profiling and performance analysis tools