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
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
Skills Required
- 8+ years experience in distributed systems, ML infrastructure, or high-performance computing
- Production-grade expertise in Python
- Low-level performance mastery: CUDA, cuDNN, Triton, kernel optimization, CPU-GPU interactions, data movement
- Experience with PyTorch and scaling training using data, model, pipeline, and context parallelism
- Design and optimize distributed training systems for multi-node GPU clusters
- Develop monitoring and debugging tools for large-scale training runs
- System-level mindset tuning hardware-software interactions for high utilization
What We Do
Genesis AI is a global full-stack robotics company developing general-purpose robots with human-level intelligence and capabilities. It aims to build foundational AI models that automate repetitive tasks across applications such as lab work and housekeeping. The company uses a proprietary physics engine to generate synthetic physical-world data, helping train robotics models for diverse real-world environments, and operates across Paris and Silicon Valley.
