Dec 2024 – Present · 1 yr 5 mos · India · On-site

• Leading automation and orchestration initiatives for AI accelerators and next-generation DPU platforms, enabling scalable validation across complex PCIe, RDMA, and multi-host topologies.
• Designed a full-stack orchestration system for Pollara AI cards, covering host provisioning, firmware workflows, virtualized environments, SR-IOV enablement, GPU passthrough, and continuous validation.
• Architected and maintained CI/CD pipelines that integrate infrastructure bring-up, Ansible workflows, and automated test execution for large regression suites.
• Expanded the test ecosystem by developing high-coverage automation across Pollara, Volcano, Salina, and Elba platforms, validating datapath behavior, firmware durability, performance, RCCL/IB interoperability, and platform functionality.
• Deep involvement in advanced I/O subsystems, including PCIe routing, ATS/ACS behavior, peer-memory models, and GPU/NIC/accelerator data paths.
• Actively researching UALink, HPC fabrics, and heterogeneous GPU/CPU/NIC interconnects to inform future cluster-level architecture decisions.

Jul 2023 – Dec 2024 · 1 yr 5 mos · India · On-site

• Built a comprehensive library of automation playbooks spanning firmware lifecycle management, network OS deployment, hypervisor provisioning, virtualization workflows, and multi-device orchestration.
• Enhanced AMD’s internal pytest-based automation framework by engineering new subsystems for device communication, host virtualization objects, and DPU/AI accelerator abstractions.
• Delivered high-volume automated validation for DPU and AI silicon, significantly increasing testing throughput and coverage across multiple generations of hardware.
• Developed CI pipelines that integrated environment setup, dependency orchestration, and automated regression execution for complex networking and AI compute stacks.
• Strengthened system expertise across Linux networking, large-scale virtualization, containerized environments, IPv4/IPv6 architectures, tunneling protocols, and modern datacenter networking practices.

Jan 2023 – Jul 2023 · 6 mos · India · On-site

• Created 40+ Ansible playbooks for firmware upgrade pipelines, SONiC OS installation, vCenter/ESXi deployment, NIC SR-IOV enablement, GPU passthrough, and Cloud Hypervisor VM orchestration.
• Designed core libraries for the internal pytest-based test automation tool, adding Telnet, Paramiko, ESXi object, and Pollara object support.
• Automated 300+ tests covering DPU and AI chip families: Pollara, Salina, Elba.
• Wrote multiple Jenkins pipelines for infra bring-up and automated regressions.
• Strong hands-on experience with IPv4/IPv6, VLAN, VXLAN, routing, NAT, virtualization, Docker, and Linux internals.