AI Factory Network Architect
Runware · United Kingdom
Infrastructureremotelead
At a glance
Highlights
- Own the network turning containerized 1MW pods into one global supercomputer.
- Build first fabrics hands-on then set standards for tens of thousands of GPUs.
- Make pod failures invisible to developers via automatic traffic rerouting.
What they offer
- Generous paid time off: vacation, sick days, public holidays
- Meaningful stock options
- Flexible hours outside core collaboration blocks
- Paid maternity, paternity and caregiver leave
- Twice-yearly company retreats
Ideal candidate
Veteran builder-operator of large-scale HPC/AI fabrics who thrives hands-on, from 3 a.m. debugging to setting standards. You want to own a global supercomputer network end-to-end and grow the team running it.
Requirements
- 15+ years designing, building, and operating high-performance networks, including HPC or AI supercomputers at thousands of GPUs scale
- Deep, hands-on InfiniBand experience: fabric design, subnet management, routing, adaptive routing, congestion control and port-level debugging
- Working knowledge of NVLink and NVSwitch domains and how scale-up and scale-out fabrics interact for multi-GPU inference and training
- remote policy
- remote-first
Lab profile
About Runware
High-performance GPU infrastructure and API for fast, scalable AI inference across models.
What they build
- Model API for image, video and LLM inference
- Serverless GPU platform for deploying AI models
- Dedicated inference capacity
Known for
- Fast real-time inference at scale
- Single flexible API across thousands of models
- Production-grade reliability and performance
- Lower cost and operational simplicity
Posted · Last seen