Gimlet
Network Engineer
San Francisco, CA. Posted 5 months ago
The role
You will build the network infrastructure connecting compute, storage, and distributed systems. You will work across network architecture, physical infrastructure, provisioning, deployment, and production operations.
What you will do
- Design and deploy network infrastructure for new data centers and hardware expansions
- Improve network performance and reliability for distributed AI workloads
- Automate provisioning, configuration, validation, and monitoring workflows
- Establish repeatable network standards
- Troubleshoot complex issues across network, hardware, and distributed systems boundaries
Required qualifications
- Experience designing, deploying, and operating production data center networks
- Strong fundamentals in routing, switching, performance, and reliability
- Experience with BGP, ECMP, EVPN, VXLAN, and spine-leaf or Clos fabrics
- Knowledge of high-performance networking technologies such as RDMA, RoCE, InfiniBand, or lossless Ethernet
- The ability to troubleshoot across network, hardware, software, and distributed systems boundaries
- Experience automating network infrastructure with Python, Ansible, Terraform, or similar tools
Nice to have
- Experience with AI/HPC, GPU, or large-scale distributed infrastructure
- Knowledge of AI application traffic patterns, including collective operations and workload colocation strategies that optimize network performance
- Experience with cloud networking on GCP, AWS, Azure, or similar platforms
- Experience with Arista, Cisco, Juniper, or NVIDIA networking platforms, as well as Palo Alto Networks PAN-OS
- Experience with network automation using Python, Ansible, Terraform, or similar tooling
- Familiarity with RDMA, RoCE, InfiniBand, or other high-performance networking environments
About the company
Gimlet is building the first multi-silicon neocloud designed for fast, efficient AI inference. They combine large-scale compute infrastructure with an execution platform that partitions AI workloads and maps each stage to the hardware best suited to run it.