Staff AI Scheduling & Orchestration Engineer
Core
Lead workload placement logic for the AI-native NeoCloud platform to eliminate GPU stranding and maximize utilization across expensive compute fleets.
Role type
Staff IC distributed systems engineer (Kubernetes scheduling)
Builds
High-performance scheduling fabric for massive-scale distributed AI training and inference jobs
Domain
AI infrastructure / High-Performance Computing (HPC)
Deliverable
production ML models
Required skills
Kubernetes scheduling frameworks, Volcano, YuniKorn, Kueue, Kubernetes Dynamic Resource Allocation (DRA), custom scheduler plugins, topology-aware pod placement, GPU hardware architectures, distributed training frameworks (PyTorch Distributed, Ray, MPI), infrastructure automation, infrastructure-as-code (Terraform, Go-based Operators)
Preferred skills
None stated
Technologies
Kubernetes, Volcano, YuniKorn, Kueue, NVLink, InfiniBand, PyTorch Distributed, Ray, MPI, Terraform, Go
Responsibilities
Design and implement advanced batch scheduling architectures for multi-node gang scheduling; Develop and manage cluster-wide admission control and job queueing mechanisms; Leverage Kubernetes DRA and custom scheduler plugins for accelerator requests; Architect topology-aware pod placement strategies; Implement automated GPU sharing technologies and multi-tenancy isolation policies; Collaborate with GPU Systems and Storage teams; Drive reliability and scalability of the scheduling stack; Mentor junior engineers and conduct design reviews
Seniority
Staff, hands-on IC with mentorship