AI Architect (內湖瑞光總部)
Core
Design and own the end-to-end technical architecture for Advantech's enterprise 'AI Factory' platform, enabling scalable, secure, and cost-effective AI agents and applications across all business domains.
Role type
Senior individual-contributor AI Architect
Builds
Enterprise AI agents, applications, model platforms, skill/MCP ecosystems, and data/RAG pipelines
Domain
Enterprise AI, Agentic Systems, Cloud Infrastructure, Data Governance
Deliverable
production ML models | product features | infrastructure
Required skills
Enterprise AI platform architecture, LLM and agentic system design, model evaluation and benchmarking, fine-tuning (SFT/LoRA/RLHF), vector stores and RAG pipelines, system integration patterns, security and compliance standards, CI/CD for AI agents, technical standards definition, hands-on prototyping and code review
Preferred skills
Internal AI platform construction, NVIDIA AI stack and on-prem GPU infrastructure, open-source LLM serving, enterprise system integration (CRM/ERP), cloud AI platform expertise (Azure/AWS)
Technologies
NemoClaw, MCP (Model Context Protocol), Azure, AWS, NVIDIA AI stack, vector stores, LLMs
Responsibilities
Define end-to-end platform architecture for AI agents and applications; Architect the NemoClaw model platform including inference optimization; Design the skill/MCP platform and agent orchestration; Establish data platform architecture for RAG and knowledge bases; Set security, guardrails, and compliance standards; Design secure system integration patterns; Define agent DevOps, observability, and cost practices; Codify architecture patterns and mentor engineers; Evaluate frontier technologies and make build-vs-buy recommendations; Prototype and review code alongside engineering teams
Seniority
Senior, hands-on IC
