Senior Software Engineer (AI Workload Orchestration)
Posted Updated
- As a Senior Software Engineer II (IC4) on the AI Workload Orchestration Platform team, you will help build and operate CoreWeave’s Kubernetes-native platform for admitting, scheduling, and operating AI workloads at scale
- This platform integrates multiple orchestration and scheduling frameworks such as Kueue, Volcano, and Ray to support modern AI training and inference workflows. It complements SUNK (Slurm on Kubernetes) by providing a Kubernetes-first, cloud-native orchestration layer with deep platform integration
- You will own meaningful components of the platform, drive reliability and performance improvements, and help scale the system as customer demand and workload complexity continue to grow
- Design, build, and operate Kubernetes-native services for AI workload orchestration and scheduling
- Own one or more platform components end-to-end, including design, implementation, testing, and on-call support
- Improve scheduling latency, cluster utilization, and workload reliability through metrics-driven engineering
- Contribute to architectural discussions across services and influence design decisions within the platform
- Work closely with adjacent teams (CKS, infrastructure, managed inference) to ensure clean interfaces and integrations
- Mentor junior engineers and raise the quality bar for code, design, and operations
- Solid understanding of Kubernetes fundamentals, APIs, controllers, and operating services in production
- Strong experience building production systems in Go (Python or C++ a plus)
- Bachelor’s Degree in Computer Science, Engineering or related technical field
- Comfortable owning services in production and participating in on-call rotations
- Experience working with scheduling, resource management, or quota-based systems
- 5–8 years of professional software engineering experience in distributed systems, cloud infrastructure, or platform engineering
- Proven ability to improve system reliability and performance using data and operational metrics
- Experience with Kubernetes-native orchestration frameworks such as Kueue, Volcano, Ray, Kubeflow, or Argo Workflows
- Familiarity with GPU-based workloads, ML training, or inference pipelines
- Experience with reliability practices including SLOs, alerting, and incident response
- Knowledge of scheduling concepts such as quota enforcement, pre-emption, and backfilling
- Exposure to AI infrastructure, HPC, or large-scale distributed compute environments