Sr. Solutions Engineer

Rafay seeks a customer-focused, technically skilled Senior Solutions Engineer.. The role involves partnering with enterprise customers/Neo clouds to architect, deploy, and optimize cloud-native infrastructure using the Rafay platform and Kubernetes ecosystems, including GPU-accelerated environments for AI/ML and inference workloads., * 8+ years of experience in customer-facing technical roles, including implementation or consulting \n

• Hands-on AI/ML engineering experience with model inference and deployment workflows \n

• Hands-on experience with NVIDIA GPU infrastructure (H100/H200/B200 preferred) \n

• Familiarity with GPU Operator, MIG, SR-IOV, and GPU fabric/networking fundamentals \n

• Experience with LLM inference serving frameworks such as vLLM, TGI, or Triton \n

• Strong Kubernetes and cloud-native environment expertise \n

• Excellent written and verbal communication abilities \n

• Deep troubleshooting skills across networking, virtualization, container orchestration, and public cloud platforms \n

• Infrastructure automation and IaC (Terraform preferred) experience desired \n

• Linux systems administration and distributed systems knowledge \n

• Proven ability to lead technical projects and manage multiple engagements \n

• Bachelor’s degree in Computer Science, Engineering, IT, or equivalent experience \n, * Experience with Run:AI and Slurm for GPU scheduling \n

• GPU autoscaling and multi-tenant GPU isolation expertise

• Deploy and configure GPU-accelerated Kubernetes clusters for AI/ML workloads, including GPU Operator, MIG, and SR-IOV setup \n

• Implement and validate LLM inference serving stacks (vLLM, TGI, Triton) during customer implementations \n

• Troubleshoot GPU cluster health, utilization, and GPU fabric/networking issues (NCCL, RoCEv2) \n

• Configure GPU observability using DCGM, OpenTelemetry, and related monitoring tooling \n

• Advise customers on distributed training and inference topology options for NVIDIA GPU infrastructure (H100/H200/B200) \n

\n

\n