Shanghai, China
AI Strategy Consultant at Veeva Systems. Sr Associate Business Analyst at Amgen （2024-2026）. MS in Analytics at USC (University of Southern California). BS in Accounting and Supply chain management with a minor in Economics at Syracuse University.
Act as a hybrid Forward Deployment Engineer and Business Consultant for AI products, partnering with clients to uncover business needs and pain points, translate them into AI use cases, and connect them with product capabilities and implementation plans.
Design Intelligence Platform Project – The Center for Design & Analysis (Research & Development) - Contributed to the development of an internal GenAI chatbot for R&D teams to address information overload in lengthy and complex clinical trial documents, enabling cross-trial comparison, study-level summarization, and trial-specific question answering. - Led output quality evaluation by defining reference-based assessment criteria and building automated evaluation workflows (e.g., RAGAS with curated ground-truth datasets), systematically analyzing response failure patterns and identifying accuracy and consistency issues related to retrieval coverage, ranking, and grounding. - Improved response accuracy by optimizing the existing RAG pipeline, which was initially implemented with LangChain for text extraction, and later enhanced by migrating retrieval orchestration to LlamaIndex and introducing hybrid search (BM25 + vector retrieval) and ranking adjustments, increasing overall retrieval accuracy to 98%+ for comparison, summarization, and Q&A tasks. Study MDM (Master Data Management) Project - Contributed to the development of an AI-assisted study data matching system within Enterprise Master Data Management, addressing large-scale, multi-attribute matching scenarios by designing a multi-agent reasoning architecture that decomposed matching logic into domain-specific agents to improve modularity and scalability. - Introduced a validation agent to reconcile intermediate matching results and reasoning paths across agents, enforcing consistency checks and conflict resolution, which increased overall matching accuracy from ~50% to 87% while significantly improving stability and explainability in high-volume workflows.
- Maintain the Dornsife Dialogues Tableau Dashboard to effectively visualize and analyze internal donors’ data. - Use Python web scraping to extract and gather information from various sources to support the BI team's requests. - Analyze and interpret data to provide insights and recommendations to the team, resulting in an average of 82% reduction in irrelevant information and an increase in accuracy and efficiency in data-driven decision making.
- Support the VP of Research & Analytics on custom market research studies and data analysis, including but not limited to distilling creative campaign insights, understanding the audience, tapping into the positioning, monitoring real-time audience reactions through social listening, and uncovering cultural trends, and segmentations. The role will primarily focus on quantitative methods. - Support project management including but not limited to: database maintenance, quality assurance, survey design / questionnaire writing, social media listening, analyzing and reporting. - Keep abreast of emerging consumer behaviors, technologies, and cultural trends that are impacting the entertainment industry. - Stay engrossed in film and series trends, box office / performance metrics, and media trends and innovations and use them to inform insights.
- Enhance user experience by implementing reporting charts, reducing master data reporting time by over 70%. - Leverage web scraping and OpenAI API to expedite long text data processing for physician and hospital information for the DS Assist Application. Transformed days-long tasks into minutes, generating accurate reference data to enhance master data, optimizing efficiency for the team with the help of artificial intelligence. - Contribute to a company-wide upskilling initiative on generative AI, creating summary reports to aid the adoption of GenAI technology within the 144-member EEA department and the entire DTI department.