Chaoran Jiang turns messy questions into things other people can use. He is finishing a B.A. in Data Science at the University of Wisconsin-Madison and spent this summer as an equity research intern at Greenwoods Asset Management in Shanghai, sizing the global AI-video market from the ground up and covering AI application equities across three markets.
The pattern repeats. At Greenwoods, the market-sizing report ships in two layers: a clean front-facing document, and an Excel backend where every figure links to the source it came from, so any number in it survives the question "where did this come from". Penetration rates are labelled as forward-looking assumptions rather than buried in a formula. The sector work runs on a self-built data pipeline behind a 33-ticker dashboard, and the 7-step investment decision framework he co-developed was reviewed by a senior buy-side investor.
At UW-Madison the same instinct became UW Course Planner: upload a DARS degree audit, pick a preference profile, get three or fewer ranked, conflict-free semester schedules with the trade-off behind every course and professor choice written out. Three unrelated data sources sit underneath it in one model. He runs the build as product owner with AI agents as the engineering team: he owns spec, roadmap and acceptance, implementation is delegated.
He started early. At 15 he founded Ignite 3D Hologram Co., designed a 3D hologram device, secured a national utility-model patent and a design patent, and pitched the business plan to two venture investors before he ever sat in a college lecture. In 2024 he cleaned finance datasets in Python at xMEMS in Santa Clara, reconciled the books in Excel and helped build the Tableau dashboard the operations team read every week.
He uses AI for leverage everywhere and owns every judgment personally. That distinction is the whole job.
Off-hours: six years of vocal training (top four in a university competition), five years of data-driven fitness training, badminton, poker.
蒋超然习惯把一团乱麻的问题，变成别人能直接用的东西。他在威斯康星大学麦迪逊分校读数据科学，明年毕业；今年夏天在上海的景林资产做投研实习，从头算了一遍全球 AI 视频的市场规模，也覆盖了美股、港股、A 股三个市场的 AI 应用类公司。
他做事的路子基本是同一套。景林那份市场规模报告分两层交付：前台是一份干净的报告，后台是一张 Excel，每个数字都能点回它的来源，被问「这个数哪来的」时不会卡壳。渗透率这类前瞻假设单独标出来，不藏进公式里。行业覆盖跑在他自己搭的行情数据管线上，33 只票一张看板；那套七步投资决策框架是他参与搭的，过了一位资深买方投资人的评审。
在麦迪逊，同样的路子变成了 UW Course Planner：上传 DARS 学位审计，选一个偏好，拿到三份以内排好序、互不冲突的学期课表，每门课、每位教授为什么这么排都写清楚。底下是三个互不相干的数据源，拼成一套数据模型。这个项目他自己当产品负责人，工程交给 AI agent 团队：spec、roadmap、验收归他，实现交出去。
他起步早。15 岁创办 Ignite 3D 全息，设计了一台 3D 全息设备，拿下一项国家实用新型专利和一项外观设计专利，还没上大学就把商业计划讲给了两位风险投资人，并按他们的意见改过一轮。2024 年在圣克拉拉的 xMEMS，他用 Python 清洗财务数据、在 Excel 里对账，还参与做了运营和销售每周看的那张 Tableau 看板。
AI 他到处用，判断全部自己负责。这个区别，就是这份工作本身。
工作之外：声乐练了六年（校级比赛前四）、按数据练了五年健身、羽毛球、德州扑克。