About this role
The role involves architecting and implementing foundational systems for next-generation AI agents, focusing on scalable long-term memory, knowledge graph pipelines, and robust context management infrastructure. This includes translating advanced research concepts like Context Graphs and GraphRAG into production-grade, reliable, and stateful AI systems.
Candidates must possess a PhD or Master's degree in a relevant field along with at least 10 years of industry experience, demonstrating deep expertise in long-term memory, knowledge graphs, or context management. Experience in top-tier AI labs or fintech/crypto environments is advantageous.
What they're looking for
AI InfrastructureAgent Memory ArchitectureKnowledge Graph InfrastructureContext ManagementGraphRAGLLMSystem DesignProduction Engineering
Frequently asked questions
What does a Principal Engineer, Agent Infrastructure & Memory Architecture at OKX do?
The role involves architecting and implementing foundational systems for next-generation AI agents, focusing on scalable long-term memory, knowledge graph pipelines, and robust context management infrastructure. This includes translating advanced research concepts like Context Graphs and GraphRAG in…
What skills does this Principal Engineer, Agent Infrastructure & Memory Architecture role need?
Key skills for this role include AI Infrastructure, Agent Memory Architecture, Knowledge Graph Infrastructure, Context Management, GraphRAG, LLM.
How much does a Principal Engineer, Agent Infrastructure & Memory Architecture at OKX pay?
The employer did not list a salary for this role. Most similar Singapore roles publish their band on the job page.
Is this Principal Engineer, Agent Infrastructure & Memory Architecture role remote, hybrid, or on-site?
This role is on-site, based in Hong Kong Island.
How do I apply for this Principal Engineer, Agent Infrastructure & Memory Architecture role?
You can apply directly on OKX's careers page. ApplyLah can tailor your résumé and cover letter to this exact role in seconds first.