Triton Software Engineer — AI Frameworks & Performance

Graphcore; West of England; On-site; GBP 65,000 - 90,000; Full time

Benefits offered by this job

Flexible work

Generous leave

Pension plan

Phantom equity

Office spaces

Free food

Life insurance

Private medical

Dental coverage

Cycle to work

Health cash plan

Job summary

Graphcore in Bristol is seeking a Software Engineer for the Triton team. You will design and optimize the framework support that enables AI developers to leverage Graphcore hardware with Triton and PyTorch.

Reporting to the Frameworks team, you will deliver new features, perform code reviews, and author technical documentation. You’ll coordinate multi-team tasks and gain deep technical ownership as the ML software landscape evolves.

Qualifications

• Proficiency in Python and C++ for building framework components.
• Strong software engineering practices, including testing and debugging.
• Experience coordinating work across multiple teams and milestones.
• Excellent written and verbal communication skills for cross-functional collaboration.
• Interest in AI, ML frameworks, and performance optimization.

Responsibilities

• Design, implement, optimize, and support core Triton framework components.
• Deliver new features and conduct code reviews within the Triton team.
• Write technical documentation and collaborate with upstream communities.
• Coordinate tasks across teams to ensure smooth integration with Graphcore hardware.
• Maintain deep technical ownership and contribute to the broader ML software stack.

Skills

Python

C++

Software engineering

Communication

Collaborative leadership

Tools

Git

Job description

Graphcore in Bristol is seeking a Software Engineer for the Triton team. You will design and optimize the framework support that enables AI developers to leverage Graphcore hardware with Triton and PyTorch.

Reporting to the Frameworks team, you will deliver new features, perform code reviews, and author technical documentation. You’ll coordinate multi-team tasks and gain deep technical ownership as the ML software landscape evolves.