04 · SELECTED WORK
从问题定义、证据采集到结构化交付，三个项目展示我如何把事情真正做出来。
From framing the question and gathering evidence to structured delivery, these projects show how I turn ideas into finished work.
work · Winnie Tang04 · 作品04 · selected work
研究、商业分析与 AI 工具并不是三条互不相关的线。它们共同体现我如何拆问题、找证据、搭结构，再把结果做成可以被使用的交付物。
Research, commercial analysis, and AI tools are not separate tracks. Together they show how I frame questions, gather evidence, build structure, and turn the result into something usable.
构建面向 A 股上市公司基金型 CVC 的三层研究数据库，把法定公告中的基金身份、LP / GP、管理人、出资、治理条款与生命周期事件转成可追溯、可复核的结构化数据。
A three-layer research database for fund-based CVC by A-share listed companies, turning statutory filings into auditable, structured data.
在 AI、半导体、新材料等赛道，从专家访谈、桌面研究和数据清洗进入问题，完成市场空间、产业链、竞争格局、商业化判断、财务预测与报告交付。
Across AI, semiconductors, and advanced materials, I worked from expert interviews and desk research through market sizing, value-chain analysis, commercialization, forecasting, and delivery.
围绕图片转可编辑 PPT、模板批量迁移、会议纪要结构化等场景，设计可复用工作流，让 AI 真正省下重复劳动的时间。
Reusable workflows for editable-slide reconstruction, template migration, and structured meeting notes—designed to save real time in repetitive knowledge work.