Performance Engineer (Junior) AI Infrastructure Cambridge (Hybrid)

• A postgraduate research background, ideally a PhD in computer science, mathematics, physics or a closely related field, strongly preferred, exceptional recent master’s graduates with directly relevant coursework will also be considered
• A genuine, demonstrated grasp of computer architecture fundamentals and how LLMs and deep learning models actually run on hardware, training versus inference, matrix multiplication, KV-caching
• Real experience building performance models or forecasting tools, Python or spreadsheet-based, from research, a thesis, a placement, or serious personal projects
• Hands-on work with GPU or accelerator code, CUDA or similar
• Familiarity with profiling tools (Nsight, PyTorch Profiler) and ideally some exposure to monitoring stacks (Prometheus, Grafana)
• Strong Python for data work, Pandas and NumPy, genuine scripting ability

Nice to have: exposure to inference serving frameworks like vLLM, published research, or open source contributions in this space.

This particular seat is for an early career professional. We’re after someone academically exceptional, ideally with a PhD, who’s ready to get stuck into real technical work quickly rather than needing a long runway to get there. Day to day You’d work alongside senior engineers on the team, pulling real metrics off live training and inference jobs and turning them into models and calculators that answer actual questions, whether an optimisation is worth shipping, whether a different setup would run cheaper. Real ownership early, with senior support close by.

Performance Engineer (Junior) AI Infrastructure Cambridge (Hybrid) up to 70k Most engineers find out whether a change works after it ships. This role is about knowing before anyone spends a penny on new hardware, building the models that predict itMy client is a Cambridge-based non-profit built on a fairly simple premise: different bits of the AI world keep solving the same infrastructure problems separately, and that’s wasteful. So they’ve built a shared space where startups, big enterprises, government bodies and university researchers can pool that hard technical work instead. Early days as an organisation, but real backing and real momentum behind it.