- Seoul, Republic of Korea
- Hybrid
- Full-time · Senior
- 8+ years
Lead AI infrastructure sales cycles in Asia Pacific and Japan, partnering with cloud service providers and sovereign AI model builders to design, benchmark and scale Intel-based AI infrastructure.
Responsibilities
- Lead architecture workshops with CTOs, platform and AI engineering teams, mapping workloads, SLAs, power and facility limits to business outcomes.
- Architect end-to-end AI infrastructure on Intel Xeon (AMX), Intel accelerators and memory (DDR5, MRDIMM, HBM, CXL), including disaggregated compute, memory and storage, prefill/decode separation, kv offload and rack-scale networking.
- Size racks, cooling (air and liquid), PUE and performance per watt, and build the TCO and carbon case for power-constrained data centers.
- Design and lead reproducible benchmarks and PoCs (LLM serving, RAG, agentic workloads) with agreed acceptance criteria, measuring tokens per second, time to first token, cost per task and performance per watt against competing platforms.
- Define reference architectures maximizing Intel goodness for agent platforms: CPU orchestration, long-context memory, sandboxes, tool and data connectors (MCP), vector databases, observability, security and data residency.
Requirements
- Bachelor's in Computer Engineering
- Bachelor's in Computer Science
- Bachelor's in Electrical Engineering
- Nice to have: CSPs and their procurement structures
- Nice to have: Intel Xeon (AMX)