AI Fabric Architect
fal · Remote
Infrastructureremotesenior
At a glance
Highlights
- Architect the network fabric powering fal's new GPU clusters.
- Own fabric from blueprint to on-floor bring-up and acceptance.
- Tune collective performance for inference and training workloads.
What they offer
- Competitive salary and equity
- Health, dental, and vision insurance (US)
- Relocation assistance to San Francisco
- Learning and growth opportunities
- Regular team events and offsites
Ideal candidate
A hands-on fabric owner who thrives on the data center floor as much as in architecture decisions. You love making large GPU fleets fast and reliable in a lean, fast-moving team.
Requirements
- 7+ years in data center or HPC networking with production GPU cluster fabrics
- Track record taking GPU fabrics from empty racks to production hands-on
- Deep expertise in InfiniBand (NDR/XDR, UFM) and/or RoCEv2 lossless Ethernet
- Strong command of rail-optimized and Clos/fat-tree designs, BGP/ECMP, EVPN/VXLAN
- Understanding of fabric tuning impact on collective communication and workload performance
- Automation with Python, Ansible or similar plus strong Linux skills
- Willingness to travel frequently to data center sites during build-outs
Lab profile
About fal
Generative media infrastructure platform for high-performance inference, orchestration, and model access at scale.
What they build
- Generative media platform
- High-performance inference APIs
- Model orchestration
Known for
- High-performance inference at scale
- Unified inference, orchestration and observability
- Idea-to-production for developers and enterprises
Posted · Last seen