VP, AI/ML Product for Frontier Research Cloud

CoreWeave; Sunnyvale (CA); On-site; USD 233,000 - 341,000; Full time

Benefits offered by this job

Medical, dental, and vision insurance

Equity awards

401(k) with match

Paid parental leave

Tuition reimbursement

Flexible PTO

Job summary

CoreWeave is seeking a visionary VP, Product AI/ML to own the product strategy and engineering for services powering AI research labs. You will bridge the metal and the researcher, delivering a high-performance cloud environment for frontier models.

Lead the Research Training Stack and orchestration tools, focusing on scalable pre-training and post-training workflows. You’ll shape roadmaps, collaborate with researchers, and drive RL/RLHF pipelines in a multi-GPU, HPC-enabled Cloud.

Qualifications

• 15+ years of engineering leadership experience.
• At least 5+ years managing large-scale infrastructure.
• Deep knowledge of Slurm and Kubernetes.
• Experience with multi-thousand GPU clusters (H100/Blackwell/Rubin).
• Background supporting frontier model research.

Responsibilities

• Oversee evolution of SUNK (Slurm on Kubernetes) for deterministic bare-metal performance.
• Drive development of next-generation orchestrators and evaluation frameworks.
• Build infrastructure for RL and RLHF pipelines.
• Partner with researchers to translate requirements into roadmaps.

Skills

Leadership

Infrastructure knowledge

Slurm

Kubernetes

Research mindset

AI/ML domain knowledge

Strategic vision

Tools

SUNK (Slurm on Kubernetes)

InfiniBand/RDMA networking

Job description

CoreWeave is seeking a visionary VP, Product AI/ML to own the product strategy and engineering for services powering AI research labs. You will bridge the metal and the researcher, delivering a high-performance cloud environment for frontier models.

Lead the Research Training Stack and orchestration tools, focusing on scalable pre-training and post-training workflows. You’ll shape roadmaps, collaborate with researchers, and drive RL/RLHF pipelines in a multi-GPU, HPC-enabled Cloud.