CUDA Engineer (AI)

Location: Remote (UK)

Company

hirify.global is a renewable energy startup building high-performance compute infrastructure to optimize power-dense GPU workloads for AI.

What you will do

• Write and optimize custom CUDA kernels for core transformer inference operations.
• Profile kernels to identify and eliminate bottlenecks in occupancy and memory throughput.
• Apply kernel fusion to reduce memory round-trips and launch overhead.
• Implement quantization-aware kernels and mixed-precision arithmetic to reduce latency.
• Build and tune caching mechanisms for efficient autoregressive decoding.
• Benchmark kernels against baselines to drive measurable performance improvements.

Requirements

• 4+ years of production CUDA experience with a track record of shipping performance-critical kernels.
• Deep understanding of GPU microarchitecture, warps, occupancy, and memory hierarchy.
• Strong CUDA C++ skills, including streams and asynchronous execution.
• Hands-on experience profiling compute-bound vs. memory-bound bottlenecks.
• Experience with kernel fusion, memory coalescing, and avoiding warp divergence.
• Solid grasp of parallel algorithm design and numerical precision tradeoffs.

Nice to have

• Experience with transformer/attention-style kernels or autoregressive decoding.
• Background in HPC or other latency-critical performance engineering.
• Exposure to multi-GPU or multi-node kernel-level optimization.
• Comfortable reading PTX/SASS to validate kernel efficiency.

Culture & Benefits

• Competitive salary and equity sign-on bonus.
• Biannual bonus scheme.
• Fully expensed tech setup.
• Breakfast and dinner allowance for office-based employees.