AI Agent & Deep Learning developer · M.Eng candidate @ Dalian University of Technology (2027) · 📍 Dalian, China
Building LLM agents, RAG pipelines & CV models with Python & PyTorch.
Just don't ask me to grind LeetCode — I'd rather ship a repo.
🤖 Agent & LLM Applications
🧠 Deep Learning & Computer Vision
🛠️ Tools
主攻 AI Agent 应用：LangGraph 编排 · RAG 检索 · 人机协同（HITL）。 Focus: AI agent applications — LangGraph orchestration, RAG grounding & human-in-the-loop.
面向电机轴加工工艺规划的生产级 Agent：输入轴几何，经 LangGraph 编排 + RAG 检索（工艺手册与案例库）生成工序路线，导出工艺卡片 Excel。
Production-grade machining process planning for motor shafts — a LangGraph workflow with RAG-grounded planning, capability filtering & rule-based verification.
- Stack：Python · FastAPI · LangGraph · ChromaDB · Jinja2
- Highlights：支持键槽/花键/齿轮/内孔/锥度；机床/刀具能力过滤；可选人机协同（HITL）决策；明确工程边界——最终参数需制造工程师复核
用 Planner + Workers 团队模式构建 AI Agent 团队：实时观看多 Agent 协作，通过链接 + 访问码发布只读、可运行版本供他人体验。
Build and share Planner + Workers AI agent teams — watch them collaborate live, then publish read-only, runnable versions via link + access code.
- Stack：TypeScript · React 19 · Cloudflare Workers + D1 (Drizzle) · OpenAI-compatible gateway
- Highlights：协作流水可见（Planning → Working → Review → Final），支持中途干预/重新规划、附件、运行历史，发布后不可变
- 🧪 prepreg-fiber-orientation — 碳纤维预浸料表面纤维取向回归（ConvNeXt-Tiny，可选 DINOv3 初始化）· CV research
- 📄 paper-figures — 面向学术论文的出版级图表工具（IEEE / Nature / Elsevier / ACM 样式预设）
P.S. I do like algorithms — I just prefer the ones that ship. 😉