Deskripsi Pekerjaan
Nexus Platform Labs is seeking a Senior AI Infrastructure Engineer to design, deploy, and scale high-throughput distributed training and inference clusters. You will partner directly with ML research and platform teams to optimize GPU cluster utilization, streamline model delivery pipelines, and enforce zero-downtime reliability across enterprise AI workloads.
Tanggung Jawab
- Architect and maintain distributed GPU infrastructure across multi-region Kubernetes clusters.
- Optimize high-throughput inference engines and reduce end-to-end token latency.
- Automate model deployment pipelines, observability, and cost allocation telemetry.
- Collaborate with ML engineers to profile kernel-level bottlenecks and improve training efficiency.
- Enforce security controls, data isolation boundaries, and failover mechanics across platforms.
- Drive incident post-mortems and automate self-healing mechanisms for cluster reliability.
Kualifikasi
- 5+ years of experience managing production-scale cloud infrastructure and Kubernetes.
- 3+ years operating large-scale distributed ML systems such as PyTorch, Ray, vLLM, or Triton.
- Strong programming proficiency in Python and Go or Rust.
- Deep expertise in GPU memory management, NCCL communication, and CUDA profiling.
- Hands-on experience with modern Infrastructure as Code using Terraform or Pulumi.
- Proven track record reducing infrastructure latency and cloud compute expenditure.