AI Data Center Infrastructure Architect

We are looking for an experienced infrastructure architect to lead the design of large-scale environments that support AI model training, inference, and high-performance computing in Irvine, California. This role blends deep technical expertise with practical decision-making across compute, networking, storage, and data center infrastructure, ensuring solutions are scalable, resilient, and ready for production. The position works closely with engineering, operations, vendors, and facilities teams to turn workload demands and site constraints into well-defined, deployable architectures.Responsibilities:• Architect end-to-end data center solutions for AI and HPC environments, aligning compute, interconnect, storage, and management systems into a cohesive platform.• Convert workload profiles, growth forecasts, and site limitations into detailed infrastructure designs that support performance, reliability, and future expansion.• Assess rack-scale platforms and hardware configurations to determine the best fit for resiliency, operational efficiency, and long-term scalability.• Design network frameworks for compute, storage, user, and administrative traffic with careful attention to bandwidth, failover, and overall fabric performance.• Define storage strategies based on throughput, latency, capacity, and application behavior to meet demanding AI and HPC requirements.• Produce detailed equipment specifications and bills of materials covering hardware components, connectivity, software, support agreements, and spare inventory.• Review vendor and integrator proposals to confirm technical completeness, pricing accuracy, and alignment with architectural requirements.• Partner with manufacturers, OEMs, and integration providers during technical assessments and contribute expertise during commercial discussions.• Establish rack layouts, cabling plans, service access considerations, and facility requirements including power, cooling, weight, and floor space.• Create validation standards for benchmarking, reliability testing, and operational acceptance before systems are released into production.