Staff ML Engineer - Developer Tooling for AI

ARM; Cambridge; Hybrid; GBP 97,000 - 132,000; Full time

Job summary

Arm’s Developer Platforms group is seeking a Staff ML Engineer to advance ML tooling and developer experiences on Arm. You will combine software engineering with ML model optimisation and deployment, building tools to analyse, optimise, and prepare models for efficient execution on Arm-based platforms.

You will influence the architecture and roadmap across ML frameworks, runtimes and hardware, delivering production-grade tooling and end-to-end workflows for developers who may not be ML experts.

Qualifications

• Strong understanding of neural-network architecture and model execution.
• Proven experience in hardware-aware model optimisation techniques (quantisation, graph optimisation, operator fusion).
• Experience with multiple ML frameworks and inference runtimes (PyTorch, ONNX Runtime, TensorFlow).
• Experience profiling ML workloads and trading off latency, memory, model size and accuracy.

Responsibilities

• Join an established engineering team to define, design and deliver production-grade model analysis and optimisation tooling for edge developers, delivered through web services and desktop applications.
• Provide technical leadership in the ML tooling space, shaping architecture and roadmap for ML developer tooling.
• Work across ML frameworks, runtimes and hardware, turning capabilities into intuitive experiences for developers.

Skills

Neural networks

Model optimisation

ML tooling

Software engineering

Cross-hardware deployment

Tools

PyTorch

ONNX/ONNX Runtime

ExecuTorch

TensorFlow/LiteRT

OpenVINO

Job description

Arm’s Developer Platforms group is seeking a Staff ML Engineer to advance ML tooling and developer experiences on Arm. You will combine software engineering with ML model optimisation and deployment, building tools to analyse, optimise, and prepare models for efficient execution on Arm-based platforms.

You will influence the architecture and roadmap across ML frameworks, runtimes and hardware, delivering production-grade tooling and end-to-end workflows for developers who may not be ML experts.