Machine Learning Engineer, Performance Tooling (AI)

Location: London, United Kingdom or Sunnyvale, California, USA

Company

hirify.global develops AI performance tooling for model training and inference across cloud and embedded hardware.

What you will do

• Design and build reliable, self-service performance tools for models, hardware targets, and development workflows.
• Define measurement standards and performance methodologies used across AI development.
• Model theoretical platform peak performance, compare it with achieved results, and identify efficiency losses at layer and operator level.
• Predict latency, memory, utilization, and compute cost before model or recipe changes consume compute resources.
• Own monitoring and regression alerting across model builds and training runs.
• Collaborate with model, compiler, runtime, platform, hardware, training, and infrastructure engineers to establish performance targets and prioritize improvements.

Requirements

• Deep hands-on performance engineering experience in complex systems, including profiling, roofline analysis, latency and throughput optimization, and root-cause analysis.
• Experience owning a tool or service end to end, from design and delivery through adoption by other teams.
• Strong Python skills and experience profiling and instrumenting large production codebases.
• Hands-on experience developing deep learning models with PyTorch.
• Ability to analyze noisy measurements and turn them into defensible conclusions.
• Strong quantitative communication and judgment for defining measurable performance questions and prioritizing impactful work.

Culture & Benefits

• Cross-functional collaboration across model, compiler, runtime, platform, hardware, training, and infrastructure teams.
• Work focused on making AI training and inference faster, more efficient, and more predictable.
• Performance decisions are supported by profiling data and quantitative analysis.