产品架构师
Core
Design the overall architecture and product roadmap for the TokenHub AI infrastructure platform, managing compute resources, model inference services, and MaaS operations.
Role type
Senior Product Architect (AI Infrastructure & MaaS)
Builds
AI compute management platforms, Kubernetes-based inference service orchestration, and MaaS product suites for enterprise customers.
Domain
AI Infrastructure, Cloud Native, MaaS, GPU Compute
Deliverable
production ML models
Required skills
AI infrastructure product design, Kubernetes/cloud-native architecture, GPU resource management, MaaS product design, inference service orchestration, multi-tenant platform design, system architecture, business scenario analysis, cross-functional collaboration
Preferred skills
vLLM/SGLang/KServe framework experience, PD separation/KV Cache optimization, GPU virtualization, FinOps for AI, hybrid cloud design, 0-to-1 platform building
Technologies
Kubernetes, GPU virtualization, API Gateway, vLLM, SGLang, KServe
Responsibilities
Define product capability layers and module boundaries for AI infrastructure; design heterogeneous compute resource pooling and scheduling; plan cloud-native inference service deployment and lifecycle management; design MaaS API access and token billing systems; optimize platform UX for administrators and developers; collaborate with R&D to productify technical capabilities; drive product commercialization and roadmap based on customer needs.
