Demonstrated expertise in AI/ML data center networking, including RoCEv2/RDMA, lossless Ethernet, and congestion control mechanisms (PFC/ECN/DCQCN), with hands-on experience validating GPU cluster fabrics, rail-optimized CLOS designs, and AI workload traffic patterns (e.g., collective communication such as NCCL). As AI and machine learning workloads reshape modern data center architectures, deep expertise in AI/ML back-end network fabrics, including RoCEv2/RDMA performance, lossless Ethernet behavior (PFC, ECN, DCQCN), and congestion management for large-scale GPU cluster environments, is required.