Data Scientist

Wenkun (Wendy) Xiao

Boston, MA +1-415-***-**** *************@*****.*** LinkedIn

LANGUAGES & TECHNICAL PROFICIENCIES

Python, SQL, OpenAI, LangChain, Azure, AWS, Databricks, Cypher, Flask, Docker, Teamcity, Snyk, Knime,PyTorch, PySpark PROFESSIONAL EXPERIENCE

Manulife - John Hancock AI Scientist Boston, MA 05/2023 - present

● Designed and deployed an end-to-end ML pipeline integrating OCR & LLM techniques to automate image information extraction and streamline insurance claims auditing, achieving 99%+ throughput, 90%+ both precision and recall, reducing erroneous file submissions by 60%+, claim drop by 8%, generating millions of dollar savings.

● Developed an LLM-empowered news monitoring tool to scrape and analyze web data for fraud risk detection.

● Developing a Multi-Agent LLM tool to enable non-tech teams to query SQL & graph databases seamlessly.

● Mining heterogeneous data to identify various fraudulent scenarios for investigation.

● Built a Conformal Prediction library tailored for diverse use cases in fraud detection with statistical inference. Aktana R&D Data Scientist San Francisco, CA 09/2019 - 10/2022

● Developed client-facing ML models to optimize lead generation and purchase propensity estimation, leveraging supervised, unsupervised & semi-supervised learning, deep learning, time series forecast, and statistical analysis:

● Designed a sales change detection system to alert sales representatives to customers’ anomalous purchase status.

● Led customer segmentation via sales allocation, LTV & customer journey analysis to drive strategic engagements.

● Developed a user engagement prediction model to enhance sales reps' interaction with the CRM platform.

● Built channel propensity models to optimize marketing efforts, improving timing, content, and channel selection. Bolt Threads Data Science Intern Emeryville, CA 11/2018 - 06/2019

● Deployed customer lifetime value (LTV) models to predict revenue and churn, supporting precision marketing.

● Designed a web scraping & NLP pipeline to perform topic extraction & sentiment analysis on competitors’ e-commerce & social media data, informing marketing strategy.

● Built personalized product recommendation engines using clustering, collaborative filtering, and predictive models. Alcon Data Analyst Singapore 11/2017 - 07/2018

● Automated workforce performance reporting with dynamic personalized targets, driving productivity improvement

● Designed database schema and operation SOP improvements, supporting the WMS database migration to SAP.

● Consolidated heterogeneous data to model the Global-Asia supply chain, optimizing regional hub locations. BABADU.ru Digital Marketing Analyst St. Petersburg, Russia 10/2016 - 07/2017

● Led international marketing campaigns by applying SEO, A/B testing, social media ad strategies, and web traffic analysis, driving 170% profit growth in the North American market. Beijing New Building Materials Group Operation Analyst Dar es Salaam, Tanzania 07/2015 - 09/2016

● Increased order-to-invoice efficiency by 3x through business workflow reorganization, SOP establishment, ERP system implementation, and automated data reporting, to support the sales expansion from wholesale to retail. EDUCATION

M.S. in Data Science, University of San Francisco 07/2018 - 06/2019 Courses: SQL, Python, R, Mathematics, Probability, Statistics, Machine Learning, Deep Learning, Experiment Design (A/B testing), Distributed Computing (Spark), Data Structures & Algorithms, Time Series, Data Visualization, NLP, etc. B.B.A, Southwestern University of Finance and Economics 09/2011 - 06/2015 Exchange study, University of California, Los Angeles 09/2014 - 12/2014 PROJECTS

Multi-Agents LLM enabled Mock Interview AI Helper

● Developed backend functions with prompt engineering for multi-agent tasking, using OpenAI API, Python, Flask, HTML. Image Classifier Web App Development, Deep Learning models implemented in Pytorch

Contact this candidate