Location: Singapore (travel-ready for regional AI cluster deployments)
Competitive Salary: From SGD $12,000
Work Hours: 9 AM – 6 PM, Monday – Friday

Location: Ubi Road (East)

Key Responsibilities

Leadership Responsibilities

• Lead multi-disciplinary engineering teams in AI cluster performance and deployment projects.

• Define technical roadmaps, standards, and best practices for large-scale AI infrastructure.

• Mentor and upskill engineers in high-performance AI frameworks, cluster optimization, and security hardening.

• Manage stakeholder communications, performance reporting, and deployment planning.

• Drive decision-making for hardware-software trade-offs, contingencies, and multi-region deployments.

• Foster a culture of reliability, proactive monitoring, and continuous performance improvement.

• Collaborate with cross-functional teams to enforce governance, Zero-Trust access, and infrastructure hardening.

Technical Responsibilities

Cluster & Hardware Optimization:

• Conduct cluster-level audits for software consistency across 3,456 GPUs and 48 racks.

• Fine-tune BIOS, firmware, kernel, and network parameters (NVLink, InfiniBand, PCIe Gen5/6) for maximum throughput.

• Validate collective communications using NVIDIA NCCL and SHARP for zero-bottleneck AI training.

AI Frameworks & Orchestration:

• Deploy, integrate, and optimize PyTorch, JAX, Slurm, Kubernetes with GPU-direct storage.

• Implement real-time telemetry for GPU/NPU health, power, and thermal metrics.

Security & Compliance:

• Implement Hardware Root-of-Trust, Secure Boot, and Zero-Trust IAM policies.

• Enforce inline encryption (AES-GCM 256) for AI fabrics and secure sensitive training data.

• Conduct vulnerability scanning, penetration testing, and ensure compliance with ISO27001, SOC2, NIST AI RMF.

• Implement secrets management for API keys, SSH keys, and SSL certificates.

Networking & Performance Tuning:

• Optimize ultra-low latency networks (400G/800G fabrics) and eliminate congestion using Sharpv4, RoCEv2.

• Audit network paths, validate topology alignment with SDN, and monitor fabric performance proactively.

• Configure and validate GPUDirect RDMA/Storage for direct GPU-to-storage data movement.

Send Full Name, Contact Number & Resume to:
📱 +65 88914996
📲 Email: [email protected]

Please include your availability, notice period and expected salary in your application.

----------------------------------------------------------------

*Only shortlisted candidates will be contacted.

CREW by HRNet | HRnet Ventures Pte Ltd (24C2435)

Athirah Bte Rosli (R2197227)