AI产品经理 AI Product Manager

2021.11 — Present

AI数据监控自动化模块落地

AI Data Monitoring Automation Module

• · 分析50余只基金运营数据，定位NAV核对占工时35%，推动AI对账模块开发
• · 设计规则引擎+机器学习+大模型三层数据监控体系，构建多Agent对账架构，设计Prompt提示词模板库保证CoT推理链条+自我校验
• · 建立A/B测试与可信度评估体系，模块上线后NAV核对工时降低35%，完成「数据→决策→验证」闭环
• · 对账准确率提升至99.7%
• · Analyzed 50+ fund operations data, identified NAV reconciliation as 35% of labor hours, drove AI reconciliation module development
• · Designed 3-layer monitoring: rule engine + ML + LLM, built multi-Agent reconciliation architecture with Prompt template library for CoT reasoning + self-verification
• · Established A/B testing and confidence evaluation; post-launch NAV reconciliation hours reduced 35%, achieving data→decision→verification loop
• · Reconciliation accuracy improved to 99.7%

RAG金融运营文档知识库设计

RAG Financial Operations Knowledge Base

• · 主导基于RAG技术的金融运营文档知识库产品设计，使用AutoGen多Agent架构(意图路由→知识检索→答案生成→事实校验)，覆盖5000余份运营文档，信息检索准确率达96%，团队工作效率提升40%
• · 设计向量数据库选型与文档Chunk切片策略，优化检索召回率
• · 构建RAG效果评测体系，使用算法指标+用户采纳率双重核心指标，建立人工+自动化双链路评测，持续驱动系统迭代
• · Led RAG-based knowledge base product design with AutoGen multi-Agent architecture (intent routing→knowledge retrieval→answer generation→fact verification), covering 5000+ operations documents, 96% retrieval accuracy, 40% team efficiency improvement
• · Designed vector database selection and document chunking strategy to optimize retrieval recall
• · Built RAG evaluation system with algorithmic metrics + user adoption rate, dual-track (manual + automated) evaluation driving continuous iteration

资管平台产品落地

Asset Management Platform Delivery

• · 负责Alpha平台(管理资产超千亿美元)新客户准入与核心模块产品化，主导多资产类别投资组合数据整合，统筹三地研发团队协作交付
• · 主导"已清算现金对账""每日NAV详细对比报告"等模块从0到1落地，通过连接不同系统数据接口产生新的报表页面，满足客户定制化需求
• · 协同美欧亚跨五个时区团队完成需求分析→方案设计→开发交付全流程，交付速度提升75%，优化数据管道架构支持日均$500M交易规模
• · Led Alpha platform ($100B+ AUM) new client onboarding and core module productization, drove multi-asset portfolio data integration, coordinated 3-region dev teams
• · Delivered "Settled Cash Reconciliation" and "Daily NAV Comparison Report" modules from 0→1, connecting cross-system data interfaces for custom reporting
• · Coordinated US/EU/Asia teams across 5 time zones, full lifecycle from requirements→design→delivery, delivery speed up 75%, data pipeline supporting $500M daily volume

高级产品分析师 Senior Product Analyst

2019.07 — 2021.10

• · 主导共同基金/私募基金需求分析，输出BRD/PRD 20+份，覆盖投资组合会计、多币种估值、TA系统及风控模块，支撑50+只基金按时交付，管理资产规模达$80B
• · 优化Geneva系统与Bloomberg/Refinitiv接口，设计数据映射与异常处理流程，数据同步准确率提升至99.5%
• · 建立数据质量监控机制，开发估值差异自动预警机制，主导3次性能压测，系统响应时间缩短40%
• · 开发的实时监控模块实现100%流程自动化
• · Led mutual/private fund requirements analysis, produced 20+ BRD/PRD covering portfolio accounting, multi-currency valuation, TA systems and risk modules, supporting 50+ fund on-time delivery, $80B AUM
• · Optimized Geneva system interfaces with Bloomberg/Refinitiv, designed data mapping and exception handling, improved data sync accuracy to 99.5%
• · Established data quality monitoring, developed valuation discrepancy auto-alert, led 3 performance stress tests, reduced system response time by 40%
• · Developed real-time monitoring module achieving 100% process automation

高级运营分析师 Senior Operations Analyst

2015.07 — 2019.06

• · 负责多资产类别交易运营管理，自主开发VBA自动化工具覆盖数据提取、差异比对、复杂佣金计算、报表生成全链路，节省1.5个FTE，人工错误率降低50%+
• · 佣金匹配时间从人工3天缩短至10分钟，年减少调节差异$1.2M，该工具被团队长期沿用成为部门标准流程
• · Managed multi-asset class trade operations, built VBA automation tools covering data extraction, variance comparison, complex commission calculation, and report generation, saving 1.5 FTE, reducing manual error rate by 50%+
• · Commission matching time reduced from 3 days to 10 minutes, annual reconciliation discrepancy reduced by $1.2M; tool became department standard process

金融学 · 硕士 MSc Finance

2013 — 2014

经济学 · 本科 BA Economics

2008 — 2012