• GPA (Core Finance Courses): 3.7/4.0 (Second-Class Scholarship)

• GPA: 93.26/100 (Ranked 1st in Major, Top 1%; National Scholarship); IELTS: 7.5 (Speaking 7.0, Fluent)

• Supported 0-to-1 payment-product planning for TikTok Shop’s launches in Brazil and Mexico. Interviewed leading competitor executives in 9+ hours of expert calls. Combined multilingual desk research with analysis of local payment flows and acquiring solutions to build a proprietary self-acquiring fee model; delivered three strategy reports informing launch decisions, driving successful launches with best-in-class cost efficiency.
• Partnered with overseas BD teams across Brazil, Mexico, the US, Japan, and Southeast Asia to optimize payment costs. Participated in and hosted 50+ cross-border meetings, and translated local payment logic, expert insights, and internal data into implementable product strategies such as order consolidation and channel switching; delivered 10+ optimization proposals that improved the average payment-cost structure by 10 bps.
• Conducted complex data analysis. Authored 40+ reusable SQL templates on TikTok’s data platform. Built dashboards for GMV, payment success rate (PSR), average order value (AOV), and other metrics.

• Built an AI bond-market monitoring system, extracting 10M+ transaction records. Applied LLMs/quantitative analysis to engineer features for 10+ investor types, developing timing signals based on institutional herding.
• Analyzed RMB 10bn+ Fixed-Income-Plus funds by scraping quarterly holdings. Applied Campisi performance attribution and multi-factor regression to isolate risks and contributions from duration, credit, and leverage.

• Led 0-to-1 development of an AI logistics rate-search agent using LangGraph to orchestrate 20+ APIs. Introduced GraphRAG to reduce hallucinations, thereby reducing enterprise customer decision time by 90%.
• Developed AI-powered CRM microservices utilizing on-device models for speaker diarization and requirement extraction. Achieved 95% accuracy in writing structured JSON data to the CRM, enhancing sales efficiency.

• Engineered an EU-compliant trading restriction API by integrating upstream execution data (pending orders, cancellation logic, anti-liquidation controls), preventing forced liquidations and ensuring compliance.
• Overhauled the compliance and risk-scoring workflow by architecting an automated AI review platform (RAG, LLM risk-scoring, three-tier automated routing), reducing review time from 1 day to 5 minutes per case.

• Independently authored the in-depth report China’s Gaming Expansion Overseas, systematically analyzing the industry value chain, competitive landscape, and core investment thesis.
• Covered key holdings including China Literature, Giant Network, and G-bits, producing 5 rapid-response research reports.
• Identified and anticipated 2 limit-up opportunities and proactively recommended trading strategies to buy-side clients.

• Led comprehensive market analysis to design a scalable dual-platform product architecture, strategically orchestrating a Multi-Agent AI system optimizing resume screening and interview preparation workflows.
• Overcame the cold-start challenge by executing a zero-budget, product-led go-to-market strategy based on competitive market analysis, successfully acquiring 1,000+ seed users and negotiating strategic B2B partnerships with three major career agencies to drive long-term sustainable revenue growth.

• Product definition and architecture: Independently planned the product from 0 to 1, defining four core modules—task management, habit formation, an AI assistant, and data insights—and designed a state-machine-driven task lifecycle (active→completed→archived) with one-click overdue postponement, subtasks, and attachments.
• AI assistant design: Designed the natural-language AI assistant “Axton” with four-way intelligent routing for task extraction, habit recognition, data analysis, and conversation; added tool-based access to user data and a long-term memory system combining user profiles with episodic memories.
• Data products and user insights: Designed a multidimensional dashboard with a 150-day activity heatmap, completion-time distribution, trend analysis, and monthly AI insights; defined completion-rate and overdue-rate metrics to close the user-feedback loop.
• Full-stack delivery: Built responsive desktop and mobile interfaces following Material Design 3 and delivered Docker-based deployment with calendar views, idea capture, full-text search, and archive management.