Senior AI Systems Engineer — Scale & Benchmarking

Graphcore; West of England; On-site; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Flexible benefits

Generous leave

Pension matching

On-site barista

Private medical insurance

Job summary

Graphcore is seeking a Senior Systems Engineer to investigate a novel AI computing architecture from rack to data centre scale. You will design experiments, benchmark AI workloads and study how hardware and software interact under real workloads, turning observations into measurable evidence.

You will help infrastructure engineers, architects and application specialists understand system behavior, performance and reliability, and you will develop tooling to collect and visualise metrics across

Qualifications

• Proficient in Python for tests, automation or data analysis.
• Strong analytical and problem-solving skills.
• Curiosity-driven approach to complex systems.
• Ability to present technical findings with reproducible evidence.
• Ability to work independently and collaborate across teams.
• Experience in AI workloads, benchmarking, systems architecture, or containers is a plus.

Responsibilities

• Design experiments to evaluate AI workloads at scale.
• Benchmark representative AI workloads and study system behavior.
• Explore how hardware and software interact across layers.
• Develop tooling and automation to collect, analyse and visualise metrics.
• Take initiative and challenge assumptions; communicate findings clearly.

Skills

Python

Analytical thinking

Curiosity-driven

Clear communication

Independent work

AI workloads / benchmarking

Job description

Graphcore is seeking a Senior Systems Engineer to investigate a novel AI computing architecture from rack to data centre scale. You will design experiments, benchmark AI workloads and study how hardware and software interact under real workloads, turning observations into measurable evidence.

You will help infrastructure engineers, architects and application specialists understand system behavior, performance and reliability, and you will develop tooling to collect and visualise metrics across