Cloud Storage Integration Engineer (Storage / Image / Registry)

Bitdeer Technologies Group • San Jose, CA

74d • $145,000 - $260,000

About The Position

Bitdeer Technologies Group is seeking an engineer to integrate distributed and parallel storage systems into its GPU cloud, supporting GPU training and inference at 10,000+ GPU, multi-region scale. The role focuses on high-throughput, low-latency storage for massive parallel I/O, including performance, multi-tenancy, and reliability. It also owns the image, driver, and registry pipeline on the node-delivery critical path as a secondary scope.

Requirements

• At least 3 years of experience in storage engineering or platform infrastructure; 6+ years for Senior-level candidates.
• Hands-on experience with distributed or parallel file systems.
• Strong understanding of distributed file system internals, including data/metadata separation, replication, consistency models, and POSIX versus object semantics.
• Proven experience integrating and operating distributed storage in production, such as Ceph, Lustre, GPFS / Spectrum Scale, BeeGFS, JuiceFS, or MinIO.
• Experience tuning performance for high-throughput, parallel I/O.
• Strong Linux systems knowledge and automation skills using Python or Go and CI/CD.

Nice To Haves

• Familiarity with NVMe, RDMA / RoCE storage networking, and caching.
• HPC / AI storage experience.
• Multi-region storage experience.

Responsibilities

• Design and integrate distributed or parallel file systems, such as Ceph, Lustre, GPFS / Spectrum Scale, BeeGFS, and JuiceFS, into the GPU cloud, optimized for AI training and inference I/O patterns.
• Own end-to-end distributed-storage integration, including provisioning, mounting, multi-tenant isolation, quotas, and lifecycle management within the platform or control plane.
• Tune storage throughput and latency for large-scale parallel access, including dataset loading and checkpointing, and benchmark across GPU SKUs and workloads.
• Architect multi-region storage, including data locality, replication and consistency, durability across failure domains, and cross-region access.
• Own golden images, templates, GPU drivers and CUDA, and the container/image registry, including versioned releases and multi-region distribution.
• Build monitoring, capacity planning, and runbooks, and eliminate single points of failure.
• Partner with Compute, Network, and Control Plane teams.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed