Principal Solutions Architect at LinkedIn
Dublin, Dublin County, Ireland -
Full Time


Start Date

Immediate

Expiry Date

08 Feb, 27

Salary

95000.0

Posted On

10 Nov, 26

Experience

20 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Yes

Skills

Industry

Information Technology & Services

Description
  • What you'll do & achieveDesign end-to-end AI/ML platform architectures across inference, training, and data pipelines
  • Build reference architectures for GPU cluster deployment, model serving, and multi-tenant ML infrastructure
  • Evaluate and recommend inference serving frameworks (vLLM, TGI, Triton, and similar)
  • Advise on GPU fabric topology for distributed training (NVLink, InfiniBand, RoCEv2)
  • Shape observability strategies across GPU metrics, OTel, eBPF, and cluster telemetry
  • Deliver technical presentations, workshops, and proof-of-concept engagements
  • Act as the primary technical advisor and escalation point for your enterprise accounts
  • Monitor and troubleshoot production: GPU utilisation, workload performance, cluster health, and cost
  • Lead root cause analysis and remediation on the hard, cross-team issues
  • Feed insight back to Product and Engineering to influence platform capability and roadmap
  • Document reference architectures and implementation guides, and mentor others as the team grows


  • Who you are8+ years in infrastructure, platform, or solutions engineering, with 3+ years focused on AI/ML infrastructure or MLOps
  • Deep Kubernetes expertise: cluster lifecycle, workloads, operators, RBAC
  • Hands-on with NVIDIA GPU infrastructure (latest-generation accelerators preferred)
  • Fluent in distributed training (NCCL, tensor and pipeline parallelism) and LLM inference serving (vLLM, TGI, and similar)
  • Familiar with GPU Operator, MIG, SR-IOV, and high-performance network fabrics
  • Strong scripting and automation skills (Python, Bash, Go preferred)
  • Comfortable across AWS, Azure, or GCP, including networking, IAM, and managed Kubernetes
  • Working knowledge of observability tooling (Prometheus, Grafana, OpenTelemetry)
  • A credible communicator who can hold their own with engineers and executives alike
  • Bonus points for Run:AI or Slurm experience, GPU scheduling and autoscaling, or certifications like CKA, CKAD, or a cloud Solutions Architect credential


How To Apply:

Incase you would like to apply to this job directly from the source, please click here

Responsibilities
Loading...