
Job Overview
Location
San Francisco
Job Type
Full-time
Category
Engineering
Date Posted
April 15, 2026
Full Job Description
đź“‹ Description
- • The DC Infrastructure Architect (Optical) at FluidStack Inc. is responsible for translating logical network topologies into scalable, resilient, and cost-optimized optical and structured cabling designs for hyperscale data centers, playing a critical role in enabling the deployment of massive compute infrastructure for artificial superintelligence.
- • Day-to-day responsibilities include breaking down DCN logical port maps into physical cross-connect and patching schedules, automating port-to-fiber mapping at scale (100K+ ports per site), designing fiber plant architecture including cassette types and trunk counts, selecting optics based on reach and loss budget, defining patching architectures, ensuring pathways scale with fiber density, establishing labeling and documentation standards, and integrating designs with DCIM tools like NetBox to track strands, ports, and patching.
- • FluidStack Inc. is a fast-moving, small team obsessed with quality, working with industry leaders like Anthropic, Google, and Meta to deploy gigawatts of compute at unprecedented speed, investing tens of billions in US infrastructure with goals of 1GW by 2026 and 10GW by 2027, where every team member owns outcomes end-to-end and treats customer problems as their own.
- • In this role, the architect will gain deep expertise in hyperscale optical network design, master structured cabling and transceiver technologies at scale, develop automation for large-scale physical layer mapping, influence the physical foundation of AI supercomputing infrastructure, and contribute to modular, future-proof designs that support transitions from 400G to 800G and beyond to 1.6T.
Skills & Technologies
About FluidStack Inc.
FluidStack Inc. operates a distributed cloud platform that aggregates under-utilized GPUs in data centers and individual machines worldwide, renting them on-demand to AI researchers, startups, and enterprises for training and inference workloads. The company automates deployment, security, and billing, offering prices up to 80% below traditional hyperscalers while providing instant access to high-end NVIDIA A100, H100, and consumer GPUs through a single API and web console. Headquartered in London, FluidStack targets machine-learning engineers who need scalable, low-cost compute without long-term commitments, claiming thousands of active nodes and customers including Fortune 500 enterprises and leading research labs.
Subscribe to the weekly newsletter for similar remote roles and curated hiring updates.
Newsletter
Weekly remote jobs and featured talent.
No spam. Only curated remote roles and product updates. You can unsubscribe anytime.
Similar Opportunities
20 hours ago
