AI Robustness Engineer vs AI Runtime Engineer

B AI Runtime Engineer Only

• Production model serving and inference pipeline architecture
• GPU/TPU resource management, scheduling, and utilization optimization
• Container orchestration with Kubernetes for AI workloads (including GPU-aware scheduling)
• Model quantization techniques (GPTQ, AWQ, GGUF, INT8/INT4) and their runtime trade-offs
• Inference framework configuration (vLLM, TensorRT-LLM, Triton Inference Server, ONNX Runtime)
• Observability and monitoring for AI services (latency, throughput, error rates, data drift, GPU metrics)
• CI/CD pipeline design for model artifacts, container images, and infrastructure-as-code
• Cost optimization and FinOps for GPU cloud spend across AWS, GCP, and Azure