Junior Performance Engineer, AI Infra - Hybrid Cambridge

Pure Resourcing Solutions Limited; Linton; Hybrid; GBP 42,000 - 70,000; Full time

Benefits offered by this job

Pension

Hybrid from Cambridge office

Job summary

Cambridge AI Infrastructure is seeking a junior Performance Engineer to join the AI infrastructure team. You will work with senior engineers to extract metrics from live training/inference jobs and translate them into predictive models and cost calculators to inform shipping decisions.

Applicants should have a postgraduate research background (PhD preferred) with strong knowledge of computer architecture, GPU programming (CUDA) and profiling/monitoring tools.

Qualifications

• Postgraduate research background, ideally a PhD, with exceptional academic record.
• Strong grasp of computer architecture and how LLMs run on hardware, training vs inference, matrix multiplication, KV-caching.
• Experience building performance models or forecasting tools in Python or spreadsheets.
• Hands-on work with GPU/accelerator code (CUDA or similar).
• Familiarity with profiling tools and monitoring stacks such as Prometheus, Grafana.
• Strong Python for data work; Pandas and NumPy; scripting ability.

Responsibilities

• Work with senior engineers to pull real metrics from live training and inference jobs and turn them into models and calculators to decide if an optimisation is worth shipping.
• Own model development early with close senior support.
• Collaborate across AI infrastructure problems to improve cost and performance decisions.

Skills

Postgraduate research background

Computer architecture fundamentals

Performance modelling / forecasting

Python for data work

GPU / accelerator code (CUDA)

Profiling tools (Nsight, PyTorch Prof)

Monitoring stacks (Prometheus, Grafana

Education

PhD in CS / Math / Physics or related field

Master's degree (relevant coursework)

Tools

CUDA

Nsight

Prometheus

Grafana

vLLM

Job description

Cambridge AI Infrastructure is seeking a junior Performance Engineer to join the AI infrastructure team. You will work with senior engineers to extract metrics from live training/inference jobs and translate them into predictive models and cost calculators to inform shipping decisions.

Applicants should have a postgraduate research background (PhD preferred) with strong knowledge of computer architecture, GPU programming (CUDA) and profiling/monitoring tools.