AI Runtime & Systems Engineer for Low-Latency Inference

Huawei Technologies Research & Development (UK) Ltd; Cambridge; On-site; GBP 65,000 - 110,000; Full time

Job summary

Huawei Research and Development UK Limited is seeking a Software Engineer to advance runtime and systems software for high-performance AI workloads, including large language models. You will optimize inference, explore graph-based execution, and collaborate with researchers and engineers across frameworks and accelerator stacks.

The role emphasizes low-latency, high-performance AI inference, with exposure to runtime systems, compilers, and multi-NPU environments in a Cambridge setting.

Qualifications

• Proficient in C/C++ and Python for AI workloads.
• Strong systems programming skills including concurrency and memory management.
• Experience with runtime or distributed AI systems is required.

Responsibilities

• Design, prototype and implement runtime optimizations for AI workloads.
• Explore scheduling, graph execution, asynchronous execution and resource management.
• Profile workloads to identify bottlenecks across CPU/accelerator stacks.
• Work with multi-NPU/accelerator platforms to optimize compute and IO.
• Collaborate with researchers and engineers across frameworks and compilers.
• Contribute to research papers and practical engineering deliverables.

Skills

C/C++

Python

Systems programming

Concurrency

Debugging

Performance analysis

Tools

CUDA

ROCm

OpenCL

CANN/Ascend

PyTorch

vLLM

SGLang

ONNX Runtime

Job description

Huawei Research and Development UK Limited is seeking a Software Engineer to advance runtime and systems software for high-performance AI workloads, including large language models. You will optimize inference, explore graph-based execution, and collaborate with researchers and engineers across frameworks and accelerator stacks.

The role emphasizes low-latency, high-performance AI inference, with exposure to runtime systems, compilers, and multi-NPU environments in a Cambridge setting.