Kubernetes调度研发工程师 - 计算
Core
Design and develop Kubernetes scheduling systems to optimize performance, resource allocation efficiency, and stability for massive clusters supporting AI workloads.
Role type
Senior IC Kubernetes Scheduler Engineer
Builds
Unified heterogeneous compute scheduling and cloud-native runtime systems for AI training, inference, and batch processing
Domain
Cloud Infrastructure / AI Systems
Deliverable
production ML models
Required skills
Go, Kubernetes Scheduler internals, batch processing scheduling, queue admission control, resource management (QoS, overcommitment), high-concurrency system design, performance profiling (pprof, tracing), cross-team system integration
Preferred skills
Experience with online/offline mixed deployment, heterogeneous resource management, fault recovery mechanisms, topology-aware scheduling
Technologies
Kubernetes, Go, cgroup, NUMA, AI training frameworks
Responsibilities
Design and optimize scheduling algorithms for large-scale clusters; Build orchestration pipelines for batch and AI jobs; Implement resource optimization features like elastic quotas and preemption; Support diverse scenarios including autonomous driving and data processing
Seniority
Senior, hands-on IC