Staff ML Engineer – Developer Tools

Overview

As Staff ML Engineer in Arm’s Developer Platforms, you will help build production-grade ML tooling that analyzes, optimizes, and prepares models for efficient execution on Arm-based hardware. You’ll shape the architecture and roadmap for ML developer tooling and ensure a seamless experience across ML frameworks, runtimes, and hardware. You work with distributed teams to deliver intuitive tools for edge developers and drive high-quality software through automated testing and strong engineering processes. This role offers the chance to influence how developers ship performant ML models on Arm from day one, in a fast-paced, collaborative environment.

Responsibilities

• Define, design and deliver edge-focused model analysis and optimization tooling via web services and desktop apps
• Provide technical leadership for ML tooling, shaping architecture and roadmap
• Collaborate across ML frameworks, runtimes and Arm hardware to create intuitive developer experiences
• Build relationships with distributed teams, product managers and UX specialists to align tooling with developer needs
• Ensure engineering quality through maintainable code, automated tests, CI, processes and customer feedback

Key requirements

• Solid understanding of neural-network architecture and model execution including computation graphs, operator dispatch, memory planning, and heterogeneous hardware execution
• Proven experience in hardware-aware model optimization techniques (quantisation, graph optimisation, operator fusion, precision reduction)
• Experience with multiple ML frameworks and runtimes (PyTorch, ONNX/ONNX Runtime, ExecuTorch, TensorFlow/LiteRT, OpenVINO)
• Experience analyzing, profiling and debugging ML workloads (latency, memory, model size, accuracy)
• Passion for building high-quality, developer-oriented tooling and workflows

• Strong collaboration across distributed teams
• Customer-focused mindset and ability to translate developer needs into tooling
• Proactive communication and stakeholder management
• Neural-network architecture and model execution concepts (graphs, dispatch, memory planning)
• Hardware-aware optimization (quantisation, graph optimisation, operator fusion, precision reduction)
• Experience across ML frameworks/runtimes: PyTorch, ONNX/ONNX Runtime, ExecuTorch, TensorFlow/LiteRT, OpenVINO

…