Training / AI Infrastructure Engineering & Research Paris

Genesis; Paris; Sur place; EUR 120 000 - 180 000; Plein temps

Résumé du poste

Genesis in Paris, France is seeking a senior ML infrastructure engineer to accelerate large-scale foundation model training by profiling and eliminating bottlenecks across the stack, from data pipelines to GPU kernels.

You will design, build, and optimize distributed training systems (PyTorch) for multi-node GPU clusters, focusing on scalability, robustness, and high utilization, with deep expertise in CUDA/cuDNN/Triton and production-grade Python.

Qualifications

• 8+ years in distributed systems, ML infra, or high-performance computing.
• Production-grade Python development experience.
• Expert in CUDA/cuDNN/Triton and kernel optimization.
• Experience with PyTorch training jobs using data, context, pipeline, and model parallelism.
• System-level mindset for hardware-software tuning and scaling.

Responsabilités

• Profile bottlenecks and reduce wall-clock time across the training stack.
• Design, build, and optimize distributed PyTorch training on multi-node GPU clusters.
• Implement low-level CUDA/CuDNN/Triton kernels and integrate into training frameworks.
• Develop monitoring and debugging tools for large-scale runs and regressions.
• Tune CPU/GPU balance, memory, data throughput, and networking for maximum utilization.

Connaissances

Python

PyTorch

CUDA/CuDNN

Distributed systems

Outils

CUDA

cuDNN

Triton

Description du poste

What You'll Do

• Drive down wall-clock time to convergence by profiling and eliminating bottlenecks across the foundation model training stack stack, from data pipelines to GPU kernels

• Design, build, and optimize distributed training systems (PyTorch) for multi-node GPU clusters, ensuring scalability, robustness, and high utilization

• Implement efficient low-level code (CUDA, cuDNN, Triton, custom kernels) and integrate it seamlessly into high-level training frameworks

• Optimize workloads for hardware efficiency: CPU/GPU compute balance, memory management, data throughput, and networking

• Develop monitoring and debugging tools for large-scale runs, enabling rapid diagnosis of performance regressions and failures

What You'll Bring

• Deep experience in distributed systems, ML infrastructure, or high-performance computing (8+ years)

• Production-grade expertise in Python

• Low-level performance mastery: CUDA/cuDNN/Triton, CPU-GPU interactions, data movement, and kernel optimization

• Scaling at the frontier: experience with PyTorch and training jobs using data, context, pipeline, and model parallelism

• System-level mindset with a track record of tuning hardware-software interactions for maximum utilization