ML Performance Engineer
Full-timeIn English
About the job
Pinely is seeking an ML Performance Engineer to join their team in Amsterdam. In this role, you will focus on optimizing model training pipelines and applying GPU-level optimizations to enhance speed and reliability.
What you will do
- Optimize model training pipeline for speed and reliability
- Apply GPU-level optimization with JAX, Triton, CUDA
- Identify and resolve performance bottlenecks in ML pipeline
- Build tools and extend internal infrastructure for training workflows
- Mentor engineers and researchers on performance best practices
- Grow team's GPU and systems-level capabilities
What you bring
- Experience optimizing neural network training in production or research
- Extensive practical experience with PyTorch or JAX
- Hands-on experience with LSTM and Transformer models
- Experience with CUDA, Triton, or low-level GPU technologies
- Proficiency in profiling tools like Nsight, cprofiler, CUDA-gdb, torch profiler
- Understanding of distributed training concepts
What you get
- High base salary and social benefits
- Generous bonus structure
- Cutting-edge hardware and software
- Ownership over business-solving initiatives
- Flexible workflow and working schedule
- Tuition reimbursement, conference and training sponsorship
Summary by Personeel.com. The full job description and the application form are on Pinely's website (apply.workable.com); Apply takes you straight there.
More jobs at Pinely
- С++ Developer (Internship)Amsterdam
- Team Lead C++ (Trading systems)Amsterdam
- Senior Research Engineer (ML infrastructure )Amsterdam
- Senior Quantitative ResearcherAmsterdam
- Senior Deep Learning ResearcherAmsterdam