I am an innovative data leader with growth mindset that incorporating AI tools into data science workflow. Specifically, my experience at Twitch includes:- Build, lead, and mentor high-performing data science teams across multiple time zones; deliver impactful insights to drive product adoption, monetization, and retention.- Lead data science for key product areas including Personalization Discovery & Recommendation, Creator Marketplace, Viewer Conversions, and Mobile Apps;- Spearheaded data science efforts for the successful launch of Twitch's Mobile App Redesign (landing page as TikTok-like short video feature and Facebook stories feature), driving critical product investment decisions and optimizations; Closely working with Ads and Commerce to trade off between user engagement and revenue leading indicators. - DS lead for several critical consumer experience products for both sides of the marketplaces including: Stream Together, Chat, Channel Points, Creator Dashboards, Milestones, Clip Creation/Export, Homepage Shelf Ranking, Enhanced Broadcast, DMCA, and more;- Foster a data-driven culture through proactive data insights sharing, rapid experimentation, product opportunity sizing, GTM strategy, and MVP evaluations

- Led data science for multiple product areas: Creator Growth products (e.g., broadcasting tools, creator dashboards & milestones), Viewer-Creator Matching via recommendations, and Twitch Mobile Apps- Drove product strategy and improvements through experimentation, in-depth analyses, and close collaboration with cross-functional partners- Partnered with product, engineering, design, and business stakeholders to define success metrics, build dashboards, and socialize insights

- Led a global data science team in Uber's Customer Obsession organization, driving measurable improvements to customer support across all Uber platforms- Conducted experiments, analyzed metrics, rolled out product optimizations, and generated actionable business insights for executive leadership- Received an Uber Business Award for an in-app help automation project that significantly reduced support contacts

- Delivered multiple high-impact analytics projects to optimize customer support, user experience, and key business metrics across Uber's products- Utilized advanced analytics techniques including experimentation, multi-touch attribution modeling, time series forecasting, and ML-based automation- Received an Uber Technology Award for a project that automated support policy decisions, saving millions annually

XQ, sponsored by Emerson Collective and Laurene Powell Jobs, is an ambitious philanthropic initiative to rethink and reshape the American high school system. In partnership with marketing and engineering teams, I am the only data analyst to provide visibility and insights to executives and key investors.• Maintain and evolve the design of database architecture and data flow.• Build, plan, launch, and testing data ETL (extract-transformation-loading) process with various teams.• Conducted real-time analyses in various stakeholder meetings.• Researched user behavior based on multiple data sources, such as Google Analytics, emails, opt-outs, call center activities, media/press reports, user website activities, CRM (customer relationship management) database, and social networks (Twitter, Facebook, Youtube, and Instagram).• Analyzed super school applications and grant proposals.• Modeled advertising effectiveness to guide media and advertising spending; provided insights for user acquisition, user experience (UX), and branding. • In charge of blog publication which communicates data-backed insights to internal and external stakeholders. • Visualized key performance metrics on dashboard (e.g., conversion rate, GIS interactive maps) for executive teams. • Highlight: created public-facing data visualization for the CEO’s presentation to political/public figures, including Joe Biden (Vice President), Yo-Yo Ma (musician), etc.

Primary responsibilities include research design, survey design, data analysis, statistical modeling, and communicating actionable insights to business/designer/engineering colleagues.Projects highlights:• Conducted psychometric, reliability, and validity studies on millions of Amplify digital software users to better understand user behavior and attitudes.• Developed new algorithms to establish national norms and national-wide growth models for over one million student users. • Managed state-wide customer-facing data deliveries and reports.• Enabled more personalized intervention based on user behavior studies and customer segmentation. • Conducted state-wide teacher evaluation studies using multiple statistical models. • Conducted product survey research including sampling, questionnaire design, focus group, interviews, missing data imputations, and data analysis. • Led longitudinal efficacy studies for Amplify products.• Conducted standard-setting studies and workshops.• Performed database quality control and led internal R and statistical training.• Collaborated with the customer and report teams to develop and define the security rules for customers’ data. • Provided analytical support for grant applications and RFPs. • Co-authored 10+ research manuals and white papers (four as the first author).• Provide expertise in predictive modeling, structural equation modeling (SEM), factor analysis, psychometrics, sampling techniques, multilevel analysis (HLM), survey design, item response theory (IRT), experimental design, latent growth modeling, test development, program evaluation, principal component analysis (PCA), differential item functioning (DIF), linking, equating, computerized adaptive testing (CAT), diagnostic classification models (DCM), standard setting, logistic regression, multivariate analysis, missing data imputations, geographic analysis.

Conducted experimental design, survey and assessment development, data collection, data manipulation, statistical analysis, and writing for publication.Projects highlights:• Explored the effects of test taking and retrieval practice in educational settings.• Research grant: Development of the Kindergarten Cognitive and Reading Assessment Tool for iPad (K-CRATI).• Explored the relationship between working memory capacity and reading education.• Explored working memory performance under stress, cognitive appraisal, and anxiety.• Investigated methodological aspects of cognitive tasks including processing speed and intelligence.• Explored college students’ leisure activities and cognitive abilities.• Conducted a factor analysis on college students’ need for cognition using LISREL and MPLUS.• Conducted a multilevel analysis of proactive interference in the working memory tasks using HLM.

Instructor for undergraduate courses: Educational Psychology, University of Florida• Combined constructivist teaching methods and direct instruction.• Duties included but not limited to creating course materials, teaching the course, developing and leading class activities and discussions, developing and grading of quizzes and exams.

Provided technical consultation at Consumer Perception & Behavior, Unilever Food & Health Research Institute, The Netherlands• Developed and explored implicit tests of positive mood evoked by Lipton milktea using stress and emotional STROOP tasks.• Developed creativity tests to explore Lipton milk tea's cognitive effects.

At Center for Brain and Cognitive Sciences• Developed the Chinese version of the Remote Associates Test and tested the validity.• Performed an ERP study of emotion regulation and attention difficulties of different attachment styles.• Investigated brain responses to the outcomes in gambling tasks using ERP and fMRI techniques.At Social Cognition Lab:• Focused on the effects of attachment styles on self-esteem, aggression, conflict resolution, learning goals, and parenting styles in various p opulations, such as enterprisers, adolescence, and urban mothers.