Staff ML QA Engineer: Build Scalable ML Test Systems

Graphcore; Greater London; On-site; GBP 80,000 - 110,000; Full time

Benefits offered by this job

Flexible working

Generous leave

Pension matching

Phantom equity

On-site barista and free food

Income protection and life assurance

EV salary sacrifice scheme

Private medical insurance and dental

Cycle to work

Job summary

Graphcore seeks an experienced engineer to build the software that validates its ML stack. You will design production-quality test infrastructure and scalable frameworks, enabling engineers to extend and rely on testing as the ML software evolves.

You’ll architect CI/CD pipelines spanning simulators, emulators and hardware, and create ML workloads to quantify performance and accuracy, guiding tooling improvements.

Qualifications

• Strong Python skills and production-quality software design experience.
• Experience designing software architecture for large, complex codebases with testability.
• Experience building CI/CD pipelines and automated testing systems in Linux environments.
• Familiarity with C or C++ and reading low-level code.
• Ability to mentor junior engineers and influence engineering practices.
• Self-directed problem solver who identifies gaps and drives improvements.

Responsibilities

• Design production-quality test infrastructure for Graphcore’s ML software.
• Architect scalable test frameworks and CI/CD pipelines across simulators, emulators and hardware.
• Create representative ML workloads, analyze execution, and derive testing improvements.

Skills

Python

Software architecture

CI/CD pipelines

C/C++

Mentoring

Problem solving

Tools

Linux

Job description

Graphcore seeks an experienced engineer to build the software that validates its ML stack. You will design production-quality test infrastructure and scalable frameworks, enabling engineers to extend and rely on testing as the ML software evolves.

You’ll architect CI/CD pipelines spanning simulators, emulators and hardware, and create ML workloads to quantify performance and accuracy, guiding tooling improvements.