PRemoteProduction Engineer, AI Storage for HPC Clusters
Context
A client running large-scale AI training needs storage engineers to build and operate the distributed file and object storage behind its training workloads. The role covers shared researcher filesystems, training datasets, and model checkpoints, with the aim of keeping them fast, reliable, and scalable. The position is fully remote with occasional travel to the San Francisco Bay Area.
What you'll do
- Diagnose and improve end-to-end storage performance across metadata, throughput, latency, caching, and network
- Automate provisioning, scaling, upgrades, and recovery with Kubernetes and infrastructure-as-code
- Move and migrate data across clusters, regions, and clouds while preserving data integrity
- Keep storage reliable through observability, testing, safe rollouts, and incident response
Profile
- Experience building or operating production distributed storage or high-performance filesystems
- Strong Linux skills and coding in Python, Go, C++, or Rust
- Experience with Lustre, VAST, Hammerspace, OpenZFS, Pure Storage, or comparable systems is a plus
- U.S. person required