Skip to main content
  1. Home/

About

·383 words· 2 min·

I’m Liu ZhuoQi, an AI Agent Architect. I turn AI from demo to production.

I don’t wrap APIs. I design Agent systems from the ground up: memory layer architecture, multi-agent orchestration, context engineering, RAG pipelines, tool-use protocols, and evaluation frameworks. I write backends in Java, Python, and Go; frontend? React, Vue, Angular, Svelte — pick your stack; and I hold a CKS (Certified Kubernetes Security Specialist) certification. Infrastructure? I’ve run OpenStack and Ceph clusters in production. Short version — I can take a model all the way to production, solo.


Why Me
#

  • Agent Architect mindset, not a prompt engineer. I think about: memory layer design (Mem0 / Zep / Letta), inter-agent communication patterns, context window economics, and the security boundaries of tool calling.
  • Security-first engineering. CKS certified with years of infrastructure security experience. I know where prompt injection hits, how to sandbox agent tool calls, and how to design data flows under GDPR and compliance constraints.
  • Full-stack delivery. Java microservices, Python agent runtimes, Go middleware — I write all three. Frontend? React, Vue, Angular, Svelte — I adapt to your stack.

Expertise
#

Agent & AI Engineering Multi-Agent Orchestration · Memory Layer Architecture (Mem0 / Zep / Letta / MemGPT) · RAG Pipeline Design & Optimization · Context Engineering & Cache Strategy · MCP & Tool-Use Protocols · Prompt Injection Defense · Agent Evaluation · LLM Deployment & Inference Optimization

Backend & Infrastructure Java / Spring Boot · Python / FastAPI · Go · Node.js · PostgreSQL · Redis · RabbitMQ / Kafka · Docker / Kubernetes (CKS) · OpenStack · Ceph · GitHub Actions CI/CD

Frontend React · Vue · Angular · Svelte · Next.js · Nuxt · Astro · TypeScript · Tailwind CSS · UnoCSS

Security CKS Certified · Container Security · API Security · AI Security (Prompt Injection / Jailbreak / Data Exfiltration) · GDPR & Data Compliance


Featured Work#

Why LLMs Have No Memory — A cross-validated research report covering 67 primary sources: Anthropic, OpenAI, Google, and Cursor official docs; original papers by Karpathy, LeCun, and Raschka; key architectures including MemGPT, Titans, Mamba-2, and Mem0.


Contact
#

If you need Agent system integration, architecture design, or security auditing — reach out via any channel above.

Liu ZhuoQi
Author
Liu ZhuoQi
把 AI Agent 做进真实产品里。写代码,也写思考。记录 AI Agent 开发、工具工程与产品落地的实战笔记。