Sep 2009 - Sep 2014
Claude Code CLI, GitHub Copilot, Cursor, Antigravity, NotebookLM, MCP, LiteLLM (Gateway), Langfuse. Agentic UI (CopilotKit), LangGraph, Agno Framework, Google ADK.
Gemini, Claude, GPT, GLM, MiniMax, Kimi. Local/offline models (Gemma, Flash) via LM Studio, Ollama.
Microservices, Modular Monolith, Spec-driven design, TDD, Event-driven Architecture (Sync/Async).
.NET Core, Angular, React, Next.js, Node.js, Redis, Kafka, ELK, Supabase.
Azure Cloud, Heroku. Docker, CI/CD (GitHub Actions, Azure DevOps).
PostgreSQL, SQL Server, MySQL, Oracle (Execution Plan, Performance tuning).
Data-driven PM, AI-integrated DevFlow, Agile/Scrum PO role optimization.
Deep experience deploying and operating on Lark: Docs, Base, IM, Approval, Calendar, Bitable. Building internal workflows and system integrations via Lark Anycross.
N8n, Anycross — automating internal workflows and cross-system integrations.
Applied LangGraph + RAG as an AI Agent platform for personalized learning paths.
Over 10 years of experience spanning both sides — from designing large-scale SaaS and ERP architectures to building project execution processes. What sets me apart is the ability to do both well: deep enough in systems to architect them, practical enough in operations to avoid designing on paper only. In the past 2 years, I have focused on applying AI to operations, automating internal workflows, building AI-integrated DevFlows, and developing a solid understanding of embedding AI into products.
Objective: To go deeper into bringing AI genuinely into products — not just at the operational layer but directly creating value for end users. Long-term direction is to be part of building meaningful products that solve real problems, rather than doing outsourced work to spec. Always committed to continuous learning to stay relevant through each new wave of technology.