Performance Engineer

CuspAI

. Build optimization layers and profiling infrastructure for large-scale training and inference workloads .

Posted 10/23/2026full-timeCambridge • United KingdomMid-LevelSeniorWebsite

Core Competencies

Role fit

Core Competencies

Use this summary to align your resume positioning with the role.

Demonstrates expertise in performance engineering and optimization for AI workloads, with a strong focus on GPU architecture and distributed training frameworks. Proficient in utilizing profiling tools and implementing strategies to enhance model efficiency and reduce latency.

Highest-signal resume keywords

Performance EngineeringGPU Architecture ExpertiseCUDA Kernel OptimizationDistributed Training FrameworksProfiling Tools Proficiency

ATS Keywords

Tailor your resume

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills

Performance EngineeringCUDA OptimizationModel PruningQuantizationDistributed TrainingMathematical Operations OptimizationData Pipeline StreamliningScaling Laws DevelopmentNetworking Protocols KnowledgeAI Workloads Optimization

Soft Skills

Builder MindsetAutonomous Work Ability

Tools & Technologies

NVIDIA GPUNsight SystemsNsight ComputeWeights & BiasesPyTorch DistributedDeepSpeedMegatron-LMNCCL

Industry Keywords

High-Performance ComputingAI WorkloadsModel FLOP UtilizationReal-Time DashboardsCluster Uptime Monitoring

Tech Stack

Tools & technologies

PyTorch

About the role

Key responsibilities & impact

Build optimization layers and profiling infrastructure for large-scale training and inference workloads
Profile current training runs to identify Model FLOP Utilization bottlenecks
Analyze model architectures and optimize mathematical operations for NVIDIA GPU kernels and NCCL collective communication layers
Implement quantization, model pruning, and distillation to reduce latency for materials search engine APIs
Streamline data pipelines to keep high-performance accelerators supplied with data
Use observability tools to trace end-to-end training-job execution and identify processing bottlenecks or interconnect contention
Build and maintain real-time dashboards for MFU, memory bandwidth utilization, and cluster uptime
Monitor and justify AI accelerator utilization, including H100 clusters, while preventing idle time and memory leaks
Develop scaling laws and internal tools to predict compute, memory, and networking behavior as models grow
Collaborate with AI Researchers to adapt model architectures for hardware compatibility
Partner with Infrastructure to inform future compute-cluster design using performance data and profiling results
Travel regularly to other European, regional, and global locations as needed
Spend at least a few weeks in the Cambridge or London office during onboarding

Requirements

What you’ll need

3+ years of experience in performance engineering, site reliability engineering, or low-level software development focused on HPC or AI workloads
Expert-level knowledge of GPU architecture
Experience optimizing CUDA kernels or using libraries such as Triton
Deep understanding of distributed training frameworks such as PyTorch Distributed, DeepSpeed, or Megatron-LM
Knowledge of networking protocols
Proficiency with profiling tools such as Nsight Systems, Nsight Compute, or Weights & Biases
Builder mindset and ability to work autonomously in an early-stage, fast-paced startup environment
Personal interest in the potential applications of CuspAI's technology
Willingness to work in Cambridge, London, Amsterdam, or Berlin offices
Existing right to work or willingness to obtain a visa for the selected location

Benefits

Comp & perks

Equity in CuspAI
28 days holiday in Germany, Netherlands, and UK, or 21 days holiday in Japan, Singapore, and United States, in addition to local public holidays
26 weeks of parental leave for primary caregivers at full pay
12 weeks of parental leave for secondary caregivers at full pay
Professional development budget
Meaningful work advancing materials science and sustainability
Interdisciplinary teamwork with world-class researchers and engineers
Reasonable adjustments during or after the interview process