
你好,我是 Hi, I am
产品经理 Product Manager
2 年互联网产品经验 2 years in internet products
教育背景 Education
当下正在接受的系统化训练。 The systematic training I'm currently receiving.

实习经历 Experience
联合创办面向美国房产经纪人的 AI 录音笔与配套 SaaS,解决客户细节遗漏和 CRM 重复录入的问题。录音笔已在 Amazon 上架,基于 20+ 场用户访谈、用户反馈以及 MySQL / Tableau 分析,确定新手引导、功能采用与留存的优先级。
用 AI 工作流搭建 SaaS 的项目记忆体验,技术栈为 React、TypeScript、服务端 API 与 Cloudflare D1/R2。录音、转写和照片会被整理成带引用和时间戳、可审阅的 AI 简报,并配有审阅与变更时间线。开发流程自动化后,一天可以完成多轮迭代。
设计两阶段抽取与校验流程,事实经用户确认后才写入。加入可断点续跑的 AI 任务,以及针对遗漏事实、无依据引用和跨对话一致性的离线评估。
Co-founded an AI recording pen and companion SaaS for U.S. real estate professionals, addressing missed client details and repetitive CRM entry. Launched the pen on Amazon; used 20+ user interviews, feedback, and MySQL / Tableau analysis to prioritize onboarding, adoption, and retention.
Built the SaaS project-memory experience with an AI workflow, using React, TypeScript, server APIs, and Cloudflare D1/R2; turned recordings, transcripts, and photos into reviewable AI briefs with citations and timestamps. Added review and change timelines; automated development enabled multiple iterations per day.
Defined two-stage extraction and verification with user approval before saving facts; added resumable AI jobs and offline evaluation for missed facts, unsupported citations, and consistency across conversations.
独立负责 ByteHi 客服 AI 人工标注平台的设计与上线:前端由本人设计并开发,与一名全栈工程师协作完成接口与上线。平台统一了全球团队的进件、标注与质检流程,标注吞吐从每小时不足 70 单提升到 250 至 300 单。
主导客服 AI 隐私过滤服务上线:把合规要求拆解为 11 类敏感数据、4 种语言,以及实时、批处理、对话三类接口。参与定义规则与模型结合的识别方案、兜底策略和质量监控,支撑全球生产环境下可审计的脱敏。
负责 AI 自我改进平台的诊断与评估模块:与算法和后端工程师协作,覆盖证据收集、根因分析、修复建议与独立验证。内部 MVP 测试中,CSAT 问题排查耗时从 2 天至 2 周缩短到 30 分钟以内。
访谈客服运营人员后,独立上线会话自然语言检索,替代原有 21 个以上的手动筛选项。定义混合检索、相关性指标、结果保存与延迟保护,并用 SQL、Apache Hive 和看板支撑分析与产品决策。
参与客服 AI Agent 的架构设计,用自然语言 Skill 替代刚性脚本,参与定义 LangGraph 与 ReAct 的推理和工具调用方式,让客服对话更灵活。
Independently owned and launched ByteHi’s manual annotation platform for customer-service AI; designed and built the frontend, partnering with one full-stack engineer on APIs and rollout. Unified intake, annotation, and quality review for global teams, raising throughput from under 70 to 250–300 cases/hour.
Led the launch of a privacy filtering service for customer-service AI; translated regulatory requirements into 11 sensitive-data categories, 4 languages, and real-time, batch, and conversation APIs. Helped define rule- and model-based detection, fallback behavior, and quality monitoring for auditable redaction in global production.
Owned the diagnosis and evaluation module of an AI self-improvement platform; partnered with ML and backend engineers on evidence collection, root-cause analysis, proposed fixes, and independent validation. Internal MVP testing cut CSAT investigations from 2 days–2 weeks to under 30 minutes.
Interviewed support operators and independently launched natural-language search, replacing 21+ manual filters for conversation investigations; defined hybrid retrieval, relevance metrics, saved results, and latency safeguards. Used SQL, Apache Hive, and dashboards to guide analysis and product decisions.
Contributed to the architecture of a customer-service AI agent, replacing rigid scripts with natural-language skills; helped define LangGraph and ReAct reasoning and tool use for flexible support conversations.
负责开发者平台的用户活跃:设计技能竞赛、虚拟货币经济、奖励循环与成长里程碑,首页 DAU 提升 35%,两月留存提升 73%。
提出并上线玩家人格测试活动:定义需求、开发前端、追踪分享漏斗,对封面、题目与角色集做 A/B 测试,20 天参与人数超过 20,000。
端到端负责 AI 技能分享社区 Skillhub,从产品规划做到前端开发,与一名后端工程师协作。33,000+ 条 AI 能力可按游戏开发岗位检索,并配有人工审核、去重、有效性校验与使用反馈,帮助团队发现和复用可靠的 AI 工具。
独立发现 AI 知识分散的问题,负责 AI 知识平台 Aico 从策略到前端开发,与一名后端工程师协作。平台整合行业洞察、内部实践、工具、课程与专家资源,稳定日活 27,000,约占腾讯员工的三分之一。
负责腾讯学堂首页改版,面向员工学习与内容发现:定义自动打标、相关性排序与创作者数据看板,并用 SQL 做产品分析。
Owned developer-platform engagement: skill competitions, virtual-currency economy design, reward loops, and progression milestones; lifted homepage DAU by 35% and two-month retention by 73%.
Pitched and launched a player-personality quiz; defined requirements, built the frontend, and tracked the sharing funnel. A/B tested covers, questions, and character sets, reaching 20,000+ participants in 20 days.
Owned AI “Skillhub”, a skill-sharing community, end to end, from product planning to frontend development with one backend engineer; made 33,000+ AI capabilities searchable by game-developer role, adding human review, deduplication, validity checks, and usage feedback to help teams discover and reuse reliable AI tools.
Independently identified fragmented AI knowledge and owned “Aico”, an AI knowledge platform, from strategy to frontend development with one backend engineer; unified industry insights, internal practices, tools, courses, and experts for 27,000 stable daily active users, roughly one-third of Tencent employees.
Owned the Tencent Academy homepage redesign for employee learning and content discovery; defined automated tagging, relevance ranking, and a creator analytics dashboard, using SQL for product analysis.
参与消费零售领域战略尽调项目,围绕中国头部餐饮连锁(星巴克中国 / 百胜中国 / 瑞幸)执行 高管薪酬对标,系统拆解各管理层级的薪酬结构与激励机制。
整合来自年报、KJRC 备案、专家访谈等碎片化数据源,构建可比薪酬模型,验证关键投资假设,识别消费零售板块的机会点。
Contributed to a strategic due-diligence project in consumer retail, conducting executive compensation benchmarking for China’s leading F&B chains (Starbucks China / Yum China / Luckin) — systematically breaking down compensation structures and incentive mechanisms across management levels.
Synthesized fragmented data from annual reports, KJRC filings, and expert interviews to build compensation models — validating critical investment hypotheses and identifying opportunities in the consumer retail sector.
完成 15+ 次利益相关方深度访谈,挖掘消费者行为变化,直接支撑内容规划与用户互动策略。
覆盖 本地选举、科技展会、文化节庆 等关键事件的多媒体报道,确保与编辑部的叙事目标对齐;将田野观察转化为战略建议,提升本地相关性与品牌共鸣。
聚焦 中国企业的海外本地化进程——从安踏在比弗利山庄的首店落子,到洛杉矶山火与社会运动等社区议题,在华语与英语读者之间搭建桥梁。
Conducted 15+ stakeholder conversations to uncover shifts in consumer behavior, directly informing content planning and engagement strategies.
Created multimedia coverage of key events (e.g., local elections, tech expos, cultural festivals), ensuring alignment with organizational messaging goals. Converted field-level observations into strategic recommendations, boosting local relevance and brand resonance.
Focused on Chinese enterprises’ overseas localization — from Anta’s debut in Beverly Hills to community issues like the LA wildfires and social movements — bridging Chinese and English-speaking readerships.
深度拆解 20+ 头部 GaaS 竞品游戏生态,系统追踪 3,000+ 次版本迭代(含平衡性调整、DLC 发布、运营活动等),提炼出”高频小更新 + 季度大版本”的 GaaS 生命周期模型。
利用 SQL 清洗业务数据,搭建 Tableau 仪表板实时监控 DAU、留存率、转化率等核心指标,输出结构化洞察报告并在内部会议上定期汇报,获多名组长级认可。
Deep-dove into 20+ leading GaaS competitor ecosystems, systematically tracking 3,000+ version updates (balance tweaks, DLC releases, live-ops events) to distill a “frequent small updates + quarterly major releases” GaaS lifecycle model.
Used SQL to clean business data and built a Tableau dashboard to monitor DAU, retention and conversion in real time — delivering structured insight reports and regular readouts that earned recognition from multiple team leads.
独立主导 himentor 微信小程序 从 0 到 1 的开发——覆盖需求定义、交互设计、前端实现到后端联调;技术栈 微信小程序原生框架(WXML / WXSS / JS)+ Node.js 后端服务 + 云数据库,负责核心页面、用户体系与导师-学员匹配流程的全链路落地。
设计并产出品牌宣传内容体系:定义视觉语言、撰写 copywriting、产出物料套组,为小程序冷启动建立内容基础。
搭建并运营 小红书产品内容矩阵——多账号定位分层(产品号 / 导师号 / 学员故事号),制定选题节奏与 hashtag 策略,将小红书流量向小程序用户池沉淀,形成可复制的增长打法。
Independently led end-to-end development of the himentor WeChat Mini-Program — from requirement definition, interaction design, and front-end implementation to back-end integration. Tech stack: WeChat Mini-Program framework (WXML / WXSS / JS) + Node.js backend + cloud database — shipped core pages, the user account system, and the mentor-mentee matching flow.
Designed and produced the full brand content system: defined visual language, wrote copy, and delivered asset kits to bootstrap the mini-program’s cold start.
Built and operated a Xiaohongshu product matrix — multi-account segmentation (product / mentor / mentee-story), established content cadence and hashtag strategy, and funneled Xiaohongshu traffic into the mini-program user base, forming a repeatable growth playbook.
参与 2022 全球数字经济大会 的一线采访与现场报道,输出多篇稿件与现场记录,服务于新华网数字经济专题的内容池。
主持 / 协调分论坛环节并参与内容创作,配合编辑部完成主题设计、嘉宾沟通、议程把控与稿件把关,确保现场内容与传播节奏一致。
为 西藏自治区旅游发展厅 项目起草招标方案——从需求调研、竞品对标到预算结构撰写,独立交付可直接提报的商务文档。
担任新华网原创视频栏目 《健康解码》 的剪辑师,完成多期节目的后期制作(粗剪、精剪、字幕、音画同步、包装输出),并配合美术团队使用 Adobe Illustrator 产出节目视觉素材。
Reported on the 2022 Global Digital Economy Conference — conducted on-site interviews and produced multiple stories for Xinhuanet’s digital-economy coverage.
Moderated and coordinated breakout sessions while contributing to content creation — working with editorial on theme design, guest outreach, agenda control, and copy review to keep on-site content aligned with the broadcast cadence.
Drafted tender proposals for the Tibet Autonomous Region Tourism Bureau project — from needs analysis and competitive benchmarking to budget structuring — independently delivering submission-ready business documents.
Served as video editor for Xinhuanet’s original program “Health Decoding” — handled full post-production (rough cut, fine cut, subtitles, audio-video sync, packaging) across multiple episodes, and produced program visual assets with Adobe Illustrator alongside the art team.
精选项目 Selected Work
一个把「勇哥餐饮创业说」方法论工程化为 Agent Skill 的餐饮创业决策系统:包含 9 篇方法论语料、30+ 真实案例、保本线计算器、快招识别器和街景 360° 选址打分模型。
基于早期对局数据的英雄联盟胜负预测:使用 6 种机器学习模型分析 9,879 场钻石段位比赛,以 72% 准确率预测 10 分钟后胜负走向。
你正在看的这个站点。基于 Astro + Tailwind + MDX,静态部署到 GitHub Pages。
想法与笔记 Writing
为什么我选择 Astro 搭建个人站
从 Next.js、Hexo、Hugo 对比到 Astro 的最终选择,以及 Content Collections 带来的开发体验。
Hello, world
这是第一篇示例文章,介绍了本站点的写作工作流与 Markdown 支持的一些常见语法。
在做了一周"Skill作家"后,我发现AI爱看这样的文章
花两天读了100篇游戏技术文章,一篇篇喂给 skill-creator 转化成 AI 能执行的"能力包"。在第14个 Skill 处,我发现了一个让人坐不住的规律:不是所有文章都能变成好 Skill——但我们可以用"换皮测试"一招判断。