AI Research Engineer - Hardware-Efficient ML

EngineersOfAI; Bristol; On-site; GBP 90,000 - 130,000; Full time

Job summary

Graphcore in Bristol seeks a Research Engineer to push AI/ML ideas from concept to scalable implementations. You will design experiments, implement them, evaluate results and present at conferences or workshops.

You will work with researchers, silicon and software engineers to advance next-generation AI hardware, focusing on low-level optimization, model scaling and practical ML workloads.

Qualifications

• Master’s, PhD or equivalent experience in a technical discipline (Maths, Statistics, CS, Physics, Chemistry).
• Proficient Python programming in modern deep learning frameworks (PyTorch or JAX).
• Familiar with deep learning fundamentals: models, optimisation, evaluation and scaling.
• Capable of designing, executing and reporting from ML experiments.
• Lower-level programming for hardware efficiency (C++/CUDA/Triton).
• Familiar with hardware capabilities for DL: threads, caches, vector/matrix engines, data dependencies, bus widths, throttling.
• Familiar with DL software stacks – compilation, kernel fusion, XLA.

Responsibilities

• Generate AI/ML ideas, design experiments, implement them and evaluate results.
• Prepare, submit and present your work to AI conferences and workshops.
• Provide technical insight to internal teams by designing experiments and delivering clear, actionable reports.
• Collaborate with researchers, silicon and software engineers to define, build and test Graphcore’s next generation of AI hardware.

Skills

Python programming

Deep learning fundamentals

ML experiments

Lower-level programming for hardware

Education

Master’s or PhD in a technical field

Tools

PyTorch

JAX

Job description

Graphcore in Bristol seeks a Research Engineer to push AI/ML ideas from concept to scalable implementations. You will design experiments, implement them, evaluate results and present at conferences or workshops.

You will work with researchers, silicon and software engineers to advance next-generation AI hardware, focusing on low-level optimization, model scaling and practical ML workloads.