Windermere, Florida, United States
Proactive, analytical and detail-oriented professional with a strong understanding of statistical programming and modeling concepts. Earned a B.S in chemical engineering, M.S in Engineering Management, and M.S in Statistics. Demonstrated abilities to utilize programming languages including SQL, R, and Python to support analytical processes, automation and business improvements. Recognized as a trusted adviser who can present logical patterns, trends, anomalies, and opportunities by accurately querying and analyzing data. Career goal is to advance within the field of data science & statistical programming. Key highlights include: Data Science & Statistical Programming • Thrive as an AGILE team player and exceed clients' changing expectation; for instance, routinely refine client problems quantitatively and then provide sustainable solutions for future applications. • Leverage machine learning models including Logistic Regression, Generalized Linear Models with Regularization (Lasso/ Ridge), CART, Random Forests, Extreme Gradient Boosted Tree and neural networks to develop optimized credit strategy for portfolio risk/revenue management, and formulate red flag policies to deter fraudsters during loan applications. • Received spotlight award from Business Analytics Centre of Excellence (Verizon) in recognition of outstanding contributions to credit tightening and bad debt mitigation during 2019 September iconic release. • Integrate learnings and model interpretations into technical documentation, intuitive data visualization and recommendation driven dynamic reporting for end business users. • DRY practitioner with a strong ability to abstract programming challenges to higher levels, discover automation opportunities and formulate a more streamlined solution to improve process efficiency. Facility Engineering & Project Management: • Demonstrated hard work, initiative and high-quality outputs as an Intern with Disney; recognized as an asset on the team and offered a rare full-time opportunity. • Strong ability to create life cycle strategies that advocate data-driven maintenance on different facility assets, foreseeing improved asset conditions and performance that aid the decision-making process. • Performed budget spending analysis to support strategic decision and provide helpful insights into future investment patterns. TECHNICAL SKILLS: Tableau | Eclipse (BIRT) | MATLAB | SAS | Teradata | SQL | R | Shiny | Hadoop | Spark | Python | Excel | JavaScript | IBM Maximo | IBM Tririga | Vueworks GIS | Jira | Confluence | Airflow
Care Transformation and Innovation Data Science Working on clinical process optimization and patient level model.
I am a data scientist working on demand forecasting of apparel products, and integration of machine learning process within an automatic pipeline. • Apply XGBoost, lightGBM, CATboost with lag features to build shipping demand forecasting for Europe and sizing forecasting for USA/CA using Python • Monitor prediction accuracy by WMAPE and perform hyperparameter tuning/retraining with Optuna • Maintain a CI/CD sellout data pipeline and build test cases to alarm drastic changes in data • Migrate Europe forecast from AWS SageMaker to GCP Vertex Instances and refactor code to reduce ~15% runtime • Research ways of model splitting and apply prophet time series on a subset of products to reduce WMAPE by ~12% • Build an analytical chatbot using LLM to enable business planners self inquiry without writing SQL during a google x Levi’s GenAI Hackathon • Share technical findings through publishing confluence pages (e.g. a detailed guide to SSH into GCP Vertex Instance through VScode)
Provide customer churn forecast consultation to B2C consumer and business group under Finance organization. Responsibilities include building data pipelines, feature research, model implementation and story telling with data visualization and business overlay. • Apply statistical models including Logistic Regression, Generalized Linear Models with Regularization (Lasso/ Ridge), CART, Random Forests, Extreme Gradient Boosted Tree to develop production ready score cuts in consumer credit and fraud red flag policy • Wireless involuntary churn forecast using Python random forest model and Cox proportional hazard model • Build a Qlik dashboard tool with REST API to automatically update churn forecast trending vs budget to client, eliminating manual update work • 5G Home credit modeling and optimization to determine deposit schemes that maximize cash flow given variable costs and write off probability • Translate R code of Loan Loss Forecasting to Python and migrate from Unix server to GCP • Cross train other team members on scorecut modeling/generation with new EFX Fraud Superscore and Neustar score • Act as subject matter expert on probability scoring for accuracy monitoring, and Oxford Economics macroeconomic correlation analysis
Provide data science consultation to the credit risk modeling space and advocate best modeling and programming practices in Business Analytics Centre of Excellence. Responsibilities include developing data pipelines, custom credit/behavioral models, scorecard engineering & validation, what-if analyses, vendor data product evaluation and git unit testing for legal compliance. • Implement Multi Adaptive Regression Spline (MARS) model to forecast performance on variable term loans to predict exposure at default • Optimize FraudIQ score threshold of red flag policies to stop identity frauds from entering credit check and prevent losses (estimated with 700K to 1 million per month) • Transform Ignite credit bureau data assets using Impala SQL in Hadoop clusters and perform quality control & data deduplication • Initiate the use of R markdown with knitR during CECL auditing to automatically generate dynamic documents that integrates model statistics, interpretation and validation plots with inline code • Develop a new scorecut automation to generate post NITP v2 score launch adapting both volume neutral and risk neutral strategy at different time • Create new R functions for team library. Examples include size compression of big GLM model RDS objects for more efficient storage, as well as empirical probability density/ cumulative distribution functions cast on grouped data
Serve as Project Planner of Walt Disney World Integrated Facility Planning (IFP), an analytical role that combines technical, engineering and business management skills in support of the company’s strategic facilities planning projects. • Project lead for strategic facilities planning of Disney Springs and ESPN Wide World of Sports ($15M annual) • Develop BIRT user reports using SQL to query data and create custom functionalities with JavaScript under Eclipse IDE • Saved the company millions by eliminating unnecessary asset replacement activities within properties and aligning projects at close proximity to reduce logistics costs. • Planned project to transition the sport field’s metal halide to LED fixtures; after the improvements the sports operation save 648,066 kWh/year for the baseball field poles and 532,057 kWh/year for the soccer fields, equivalent to a total savings of 44% of the original energy consumption. • Tasked with creating baseline facility plans for Shanghai Disneyland defining potential capital projects for the next 10 years. • Recipient of “Disney 4-Keys” recognition for creating a project tracker with VBA Macros along with a self-contained reporting mechanism and edit synchronization feature to streamline a separate department’s processes.
As an Intern for the Integrated Facility Planning department, partnered with facility engineers to evaluate the conditions of various property assets including roofing, light poles, and HVAC systems. • Analyzed conditions and determined asset performance; predicted when the next major maintenance or upgrade would need to occur. • Collaborated with an architectural consulting firm to convert Disney roof survey results into quantifiable data, which were utilized to plot roof degradation curves for different materials. • Developed an Excel macro that automatically recalculated and plotted lifecycle costs of ownership over time based on users’ input on lifecycle activities placement. • Noteworthy project included independently creating a central repository for projectors; identified all projectors across locations (approximately 1,840) and created an inventory with comprehensive specifications and warranty documentation; this important project helped to facilitate future replacement and parts ordering, avoiding unnecessary costs.
Planned, coordinated and implemented multiple large-scale events including the Employment Opportunity Fair in Indianapolis. • Maintained databases of exhibitors, managed inventory space and designed exposition spaces (maximizing the space and accurately estimating material costs) via ExpoCAD software. • Solicited and retained corporate sponsorships to support youth educational programs.