Fluidstack
Network Engineer, Design & Engineering
$202K–$261K
New York, NY · Hybrid · Full-time · Posted Oct 30, 2026
About the job
Designs and validates data center network architectures and fabrics for AI training and inference, from customer requirements through deployment-ready specifications.
What you'll do
- Own network design lifecycle and produce topology, addressing, routing, and fabric configuration designs across front-end, back-end, and storage networks.
- Design lossless RoCEv2 Ethernet fabrics and translate logical designs into physical layouts, cabling, fiber, and equipment-placement plans.
- Create HLDs, LLDs, BOMs, cutsheets, and cabling matrices; lead design reviews and develop reusable reference architectures.
What they ask for
- Experience designing data center network fabrics from requirements through deployment.
- Deep L1–L3 expertise, including CLOS/fat-tree topologies, BGP, and EVPN/VXLAN.
- Experience designing lossless RDMA (RoCEv2) fabrics and understanding congestion management at training scale.
- Ability to reason through novel design challenges and partner across hardware, operations, ICT, software, and validation teams.
Nice to have
- Source-of-truth-driven design generation.
- Multi-vendor GPU platform integration or large-cluster (100k+ accelerator) experience.
Pay
$202K – $261K USD total pay
Tools
BGP, EVPN/VXLAN, RoCEv2, CLOS/fat-tree, ECMP
AI focus
Designs network infrastructure for AI training and inference, including large-scale GPU training fabrics.
Model-written summary · the original posting is authoritative