Machine Learning Engineer -Ethernet/IP networking, including L2/L3 forwarding, BGP, ECMP, EVPN/VXLAN, VRFs, ACLs, QoS, 5 to 8 years

Cisco — Bangalore, India

Meet the Team You will join Cisco's AI infrastructure networking team, focused on validating high-performance Ethernet fabrics for AI/ML workloads and helping customers and partners deploy scalable, resilient data center networks. Your Impact In this role, you will plan and run AI/ML infrastructure benchmarks, validate fabric designs, and translate results into reference architectures, deployment guidance, and customer-ready recommendations. The work connects lab validation with real customer outcomes across switching, host networking, RoCE fabrics, automation, and telemetry. Plan, execute, and analyze AI/ML network benchmarks covering throughput, latency, congestion, incast, failover, path diversity, and workload performance. Design, implement, and validate L3 leaf-spine fabrics for AI infrastructure using eBGP/iBGP, BGP unnumbered, EVPN/VXLAN, VRFs, ECMP, multi-tenancy, and SRv6. Define benchmark methodology, KPIs, pass/fail criteria, repeatability requirements, and reporting for AI Ethernet/RoCE fabrics. Run performance tests with NCCL, RCCL, ib_write_bw, ib_read_bw, ib_send_bw, ib_write_lat, netperf, iperf, MPI, OSU benchmarks, and microburst methods. Create validated designs, reference architectures, configuration templates, runbooks, benchmark reports, and deployment recommendations for customers and engineering stakeholders. Minimum Qualifications 4-6 years of hands-on experience in infrastructure benchmarking, network validation, data center networking, AI/cloud infrastructure, network implementation, or systems performance engineering. Hands-on experience designing, configuring, and troubleshooting L3 leaf-spine fabrics. Strong understanding of Ethernet/IP networking, including L2/L3 forwarding, BGP, ECMP, EVPN/VXLAN, VRFs, ACLs, QoS, MTU, telemetry, and troubleshooting. Working knowledge of RDMA, RoCEv2, PFC, ECN, congestion control, QoS queues, buffer behavior, and lossless Ethernet concepts. Hands-on Linux and automation experience using Python, Ansible,

View on HireSetu