Thoughts
思考
Thoughts, research, and practical experiments in data, AI, finance, and independent work.
关于数据、AI、财务与独立项目的思考、研究和实践记录。
Reasoning & Trust
推理与信任
From human worldviews, to explicit AI reasoning, to the governed data foundations enterprise intelligence requires.
从人的世界观，到显式的 AI 推理，再到企业智能所需要的受治理数据基础。
From WisdomAI to Enterprise Finance: Ontology Defines Meaning, Knowledge Graph Connects Facts
从 WisdomAI 到企业财务：Ontology 定义意义，Knowledge Graph 连接事实
How a knowledge-modeling question reveals the semantic architecture needed for governed financial analysis.
从知识建模问题出发，理解受治理的企业财务分析需要怎样的语义架构。
Enterprise Finance AI Needs More Than RAG
企业财务 AI，需要的不只是 RAG
What should remain structured, what belongs in embeddings, and how ontology gives a finance agent a governed enterprise world model.
什么必须保留为结构化事实，什么适合进入 Embedding，以及 Ontology 如何为财务 Agent 建立受治理的企业世界模型。
Same Problem, Different Worldviews
同一个问题，不同的世界观
One failed school presentation reveals why AI should make its reasoning framework explicit before giving advice.
一次失败的课堂演讲，揭示了为什么 AI 在给出建议之前，应该先明确自己的推理框架。
Why Every Decision Needs a Reasoning Framework
为什么每一个决定都需要一个推理框架
From human worldviews to an explicit, traceable architecture for AI and enterprise decisions.
从人的世界观，走向 AI 与企业决策中显式、可追溯的推理架构。
Trusted Data Is Where Enterprise AI Begins
可信数据，才是企业 AI 真正的起点
Why governance-qualified data and a governed semantic layer are the foundation for enterprise AI ontology, reasoning, and trusted action.
为什么治理合格数据与受治理的语义层，是企业构建 AI Ontology、可信推理与可控行动的起点和基石。
Products & Experiments
产品与实验
What building real AI products teaches me about knowledge, context, and trust.
真实的 AI 产品实践，让我重新理解知识、上下文与信任。
The Best Automation Knows When to Stop
好的自动化，知道什么时候该停下来
How Chloe Resume Tailor and a human-in-the-loop workflow reduced each targeted remote BA/DA application from 20–30 minutes to under five.
Chloe Resume Tailor 与 Human-in-the-loop 流程，如何把每次远程 BA/DA 定向投递从 20–30 分钟缩短到 5 分钟以内。
WisdomAI Should Chunk Meaning, Not Characters
WisdomAI 应该按义理切分，而不是按字符切分
How I am designing reasoning-aware chunking, authority layers, and traceable retrieval for the Buddhist MVP of WisdomAI.
我如何为 WisdomAI 佛教 MVP 设计推理感知切分、权威分层与可追溯检索。
Why I Stopped Building AI StudyMate
为什么我停止开发 AI StudyMate
A classroom project about learning continuity led me to design an AI tutor—and discover that trusted knowledge was the hardest part to build.
一个关于学习连续性的课堂项目，让我开始设计 AI Tutor，也让我发现：最难构建的并不是 AI，而是可信知识。