Senior AI Storage Architect for GPU-Accelerated Cloud

Bitdeer Technologies Group; Singapore; On-site; SGD 120,000 - 160,000; Full time

Benefits offered by this job

Welfare benefits

Training & mentoring

Development opportunities

Job summary

Bitdeer Technologies Group is seeking a storage-focused engineer to deploy and operate high-performance storage systems for AI workloads. You will design architectures optimized for AI data patterns, implement multi-tenant QoS, and manage storage networking for GPU clusters.

Strong Linux, benchmarking, and automation mindset are essential. In this role, you will partner with the platform team to instrument telemetry, build runbooks-as-code, and drive observability improvements.

Qualifications

• 5+ years in storage operations for enterprise or HPC environments.
• Experience supporting AI/ML workloads and high-performance storage networks.
• Experience implementing multi-tenant storage with QoS and isolation.
• Strong Linux systems knowledge including kernel tuning and filesystem internals.

Responsibilities

• Deploy and operate storage systems (WEKA, VAST Data, Ceph, DDN/Lustre) optimized for AI workloads.
• Design AI-oriented storage architectures for patterns like checkpoint I/O bursts and fast dataset reads.
• Implement multi-tenant storage isolation with QoS, quotas, and access controls; optimize GPU Direct Storage.
• Deploy and manage storage networking (NFS over RDMA, NVMe-oF) and Nvidia CMX for orchestration.
• Diagnose and tune storage performance using fio/IOR/mdtest; maintain runbooks for common failures.
• Plan capacity for GPU clusters and manage firmware, data migration, and DR procedures.
• Instrument telemetry (latency, checkpoint times, NVMe SMART) and feed to metrics/logs.
• Collaborate to define storage-fault predictor signals and labels; convert incidents into automation.
• Develop observability baselines and reduce MTTR for storage incidents.
• Ensure new Nvidia GB200-class clusters align with the storage architecture.

Skills

Storage operations

AI/ML workloads

Observability

Performance benchmarking

Multi-tenant QoS

Linux systems

Tools

WEKA

VAST Data

Ceph

DDN/Lustre

NFS over RDMA

NVMe-oF

GPU Direct Storage

fio

IOR

mdtest

Kernel tuning

Filesystem internals

Block device management

Job description

Bitdeer Technologies Group is seeking a storage-focused engineer to deploy and operate high-performance storage systems for AI workloads. You will design architectures optimized for AI data patterns, implement multi-tenant QoS, and manage storage networking for GPU clusters.

Strong Linux, benchmarking, and automation mindset are essential. In this role, you will partner with the platform team to instrument telemetry, build runbooks-as-code, and drive observability improvements.