AI Infra Engineer | System Builder | Blogger

Based in Shanghai, China 🇨🇳

I build infrastructure for AI systems — from LLM inference optimization to distributed training platforms. My work sits at the intersection of machine learning and systems engineering.

Currently focused on:

• LLM Inference — vLLM, KVCache optimization, inference serving
• AI Infrastructure — Serving systems, PD separation, throughput optimization
• Distributed Systems — Kubernetes, container orchestration, scheduling
• Systems Programming — Linux kernel, eBPF, networking

I'm a true agent lover. Recently deep-diving into Agent architecture and Agent Skills best practices:

• Building self-improving Agent Skills with feedback loops
• Designing objective metrics vs subjective evaluation for Agent outputs
• Token budget control and iteration termination strategies
• Context engineering for persistent Agent memory
• Balancing MCP tools vs scripts for Agent automation

Latest writing: Agent Skill Self-Improvement Mechanisms

I write about AI infrastructure, systems programming, and lessons learned building production ML platforms.

Read my blog →

Recent topics:

• Agent Skill self-improvement mechanisms
• LLM inference benchmarks
• vLLM architecture deep-dives
• FlashAttention analysis
• Distributed training systems

"Text > Brain" — Write things down. Memory fades, files persist.

"Objective metrics > Subjective evaluation" — Let numbers speak when they can.

"Scripts > MCP for fixed workflows" — Use the right tool for the job.

I run an AI agent called OpenClaw on an old Mac at home. It helps me with coding, writing, and automation. The agent has its own memory system, heartbeat checks, and can deploy to Vercel.

This profile was written by OpenClaw — based on my blog posts, coding patterns, and technical interests it observed over time.

Yes, my house has a resident AI now. 🤖

• Blog: ggaaooppeenngg.github.io
• GitHub: @ggaaooppeenngg
• Email: See my blog

Last updated: March 2026