SUMMARY:

Data scientist with 2+ years of work experience in data analytics and the master degree in Statistics. Proficient in working with large and complex datasets to solve analytical problems using statistical analysis, machine learning and mathematical modeling methods. Experience in optimization and NLP. Strong programming skills in Python, R and SQL.

SKILL:

Python: pandas, scikit - learn, keras, PuLP, matplotlib, folium, seaborn, NLTK, pySpark, etc.

R: data.table, dplyr, tidyr, H2O, xgboost, ggplot2, R Shiny etc.

Database: Oracle, MySQL, MS SQL Server

Software: SAS, Matlab, Google Analytics, Tabular, Domino data lab, Attensity

EXPERIENCE:

Data Scientist

Confidential, Northbrook, IL

Responsibilities:

• Preformed and led the development of Confidential tool - an internal data product to help market manages to optimize our network coverage
• Developed a demand forecasting model using space-time ARMA model, and built an optimizer that maximized the coverage of predicted demand with existing service providers by solving the linear programming problem to figure out the location/time of out-of-network ( Confidential ) services to help network team plan for the supply shortage
• Developed a machine learning model to predict travel time of the service and built a maximum coverage optimizer with travel time constraint, then applied clustering models on the Confidential services to determine the number of new service providers needed to build the network for luxury vehicles
• Built a regression model to predict the cost of Confidential services to help dispatcher negotiate prices with service providers
• Built a clustering model for GHRN rescuers segmentation and applied survival analysis to estimate rescuers’ attrition to help operation team drive strategies on right rescuers to improve their performance and reduce turnover
• Built a topic model using Latent Dirichlet Allocation (LDA) on call note text data to find reasons for calls unlinked with services to help call center improve the implementation of dispatch system and reduce call cost
• Applied text mining techniques on customer satisfaction survey data to find customer complaints topics and built an automated weekly report of topic frequency for a senior leadership meeting using Attensity
• Analyzed the impact of launching Status Texting program on key metrics (recall per rescue, NPS) using statistical techniques (hypothesis testing, regression)
• Designed metrics to measure the digital utilization and call center performance and then studied the impact of KPIs on these metrics to help business team drive strategies to improve the digital products and reduce call center cost

Business Intelligence Analyst

Confidential, Sunnyvale, CA

Responsibilities:

• Worked with project manager to understand business requirements and translated into analytical problems with SQL
• Created reports in MS SSRS using sales database to efficiently compare KPIs across time and employee; maintained and modified existing reports and dashboard to adapt to new business needs

Data Analyst

Confidential, Sunnyvale, California

Responsibilities:

• Defined the data collection needs using Google Analytics and metrics that tracked user engagement across various aspects of the mobile app to improve the efficiency of data collection process and data analysis in the future
• Analyzed the key metrics with statistical techniques (hypothesis testing, regression, etc.) to identify the KPIs that influenced the user growth of the mobile app and detected potential bugs in the program, and significantly increased the number of daily active users by 109%
• Performed data visualization by making plots with ggplot2 and building dashboard in Google Analytics to gain insight on user performance and help decision-making
• Created the weekly reports using node.js by extracting, transforming and loading data from multiple data sources (Google Analytics, back-end database, etc.)
• Designed the algorithms of pipeline from the scratch (including summarizing the content of text files, transcribing audio files to text) with Python by collaborating with engineers

Actuarial Analyst Intern

Confidential

Responsibilities:

• Retrieved data from database for different tasks with SQL queries in Microsoft Access
• Collected, cleaned and manipulated actuarial data from multiple sources to perform valuation of insurance products
• Performed data conversion from Confidential to Prophet by developing Macros in Microsoft Excel, saving 70% of the time