Principal AI and ML Infra Software Engineer, GPU Clusters
Core
Enhancing efficiency for AI/ML researchers by implementing progressions throughout the entire stack and crafting potent, effective, and scalable solutions for GPU Clusters.
Role type
Principal AI and ML Infra Software Engineer (GPU Clusters)
Builds
Potent, effective, and scalable AI/ML solutions on GPU Clusters
Domain
AI/ML infrastructure, High Performance Computing (HPC), Accelerated Computing
Deliverable
infrastructure
Required skills
HPC infrastructure operations, accelerated computing (GPU, custom silicon), storage systems (Lustre, GPFS, BeeGFS), scheduling & orchestration (Slurm, Kubernetes, LSF), high-speed networking (Infiniband, RoCE, Amazon EFA), containers (Docker, Enroot), distributed training (PyTorch DDP/FSDP, NeMo, JAX), Python, Go, Bash, parallel computing frameworks
Preferred skills
Cloud computing platforms (AWS, GCP, Azure)
Responsibilities
Engage with AI/ML research teams to discern infrastructure requirements and barriers; Proactively identify researcher efficiency bottlenecks and lead initiatives to systematically improve them; Monitor and optimize infrastructure performance ensuring high availability, scalability, and efficient resource utilization; Help define and improve measures of AI researcher efficiency; Work with researchers, data engineers, and DevOps to develop a cohesive AI/ML infrastructure ecosystem; Keep up to date with recent developments in AI/ML technologies and advocate for their integration
Seniority
Principal, hands-on IC with strategy & mentorship
