You’ll bring deep, vendor-agnostic expertise across the AI data path, with hands-on experience of one or more leading high-performance storage and memory platforms—systems in the class of WEKA, VAST Data, DDN, or Dell (PowerScale/PowerFlex)—understanding the underlying principles well enough that the specific brand is secondary. You’re fluent in NVMe and NVMe-oF, parallel and software-defined storage, object and file systems, and the low-latency network fabrics (RDMA, RoCEv2, InfiniBand) that link data to GPUs, and you track next-generation memory directions such as CXL, memory disaggregation, and KV-cache optimisation for LLM workloads.
Just as important is benchmarking discipline – you can design and run workload-driven evaluations across bare-metal, virtualised, and containerised environments and translate the numbers into architecture and business cases. You’ve done this at scale, integrating storage and memory into GPU clusters (DGX/HGX and similar) and reasoning about the CAPEX, server-count, and power trade-offs that come with it. Typically this comes with 10+ years in storage, memory, HPC, or AI infrastructure roles.
Communication rounds it out. You can advise a C-level customer, author a clear reference architecture or whitepaper, and mentor engineers with equal ease, and you’re comfortable in high-ambiguity environments where the right architecture has to be created rather than copied.