Own end-to-end reliability, performance tuning, and fault resolution across large-scale GPU computing clusters
Architect stability governance across Kubernetes, RDMA/InfiniBand fabrics, and distributed AI training environments
About the company
This fast-scaling technology firm builds foundational digital infrastructure designed to support complex, next-generation software applications. Through strategic capital investments and streamlined operations, the organization enables enterprises to deploy and run data intensive systems reliably at scale.
About the job
Owning operational stability across intelligent compute clusters encompassing GPU hardware, Kubernetes orchestration, high-speed networking (RDMA, RoCE, InfiniBand), and AIOps tooling
Leading root cause investigations and end-to-end remediation for complex failures across distributed model training, inference, and benchmarking workloads
Diagnosing deep infrastructure anomalies including GPU hardware faults, job scheduling bottlenecks, storage I/O errors, and network performance jitter
Developing emergency response protocols, fault-handling playbooks, and standardized diagnostic runbooks to accelerate incident recovery
Executing performance benchmarking and stability optimizations across multi-node training environments and GPU resource schedulers
Skills and experience required
Deep command of Linux operating system internals and low-level system, net