BI Tools: MS Office Suite, Tableau, Spotfire, QlikView, IBM Cognos, Alteryx, ClickFox Experience Analytics (CEA)
Platforms: Hadoop (Hive, MapReduce, Spark, etc.), IBM Bluemix, MS Azure, IBM Watson, AWS
SELECTED PROJECTS (Mainly used Python):
HOME DEPOT PRODUCT SEARCH RELEVANCE, KAGGLE COMPETITION New York, NY | 3/16 - 5/16.
Built a search engine that returns the most relevant products given a search term in order to improve customer experience, and ranked 303 out of 2125 teams in the competition (top 15%).
Utilized a variety of Text Mining and Data Mining techniques such as Feature Engineering, TF-IDF, N-gram, Latent Semantic Analysis, Random Forest algorithm, and K-Fold cross-validation
A BIG DATA APPROACH TO CANCER BLOG ANALYTICS, RESEARCH PROJECT New York, NY | 12/15 - 8/16.
Developed a tagging system (80% Accuracy) that can generate topic labels to a health care article using K-means Clustering model and Naïve Bayes Classification model.
Extracted key information from 2 million cancer blogs using NLTK package, and visualized the results using word clouds.
Created a lexicon that contains 500+ keywords from scratch for cancer related topics to improve accuracy of the model
MARCH MADNESS DATA CRUNCH - NCAA PREDICTION COMPETITION New York, NY | 2/16 - 3/16.
Crawled data from different websites such as AP Poll using Beautiful Soup package, Merged and cleaned dataset using Pandas and Regular Expression packages.
Built a supervised Machine Learning model using Scikit-learn package to predict wining probabilities of all teams, which used inputs such as ranking seeds, historical win/lose records, defense/offence efficiency scores, etc.
Measured performance of models using Confusion Matrix, Recall, Precision, and ROC curve.
08/2016 to Current
Analytics ConsultantATOS － Princeton, NJ
NAO Journey Analytics Practice.
Reduced the time required from Analysts to perform basic exploratory analysis from 2 days to 20 mins by developing Python applications to automate analytics processes, such as estimating potential cost savings for a specific use case, generating calculated attributes and perform deep-dive analysis.
Enabled Siemens Global and L'Oréal account teams to understand and address customer journeys leading to repeat calls, poor agent performance and synchronization issues, and recommended solutions to help them leverage journey insights to reduce costs (~$2-3M) and improve customer satisfaction (~0.1-0.4 CSAT score).
Saved 40 hours manual work for each project by leading a text mining project to analyze millions of agent notes from three ticketing systems (SDM12, ServiceNow and Remedy) and uncover reasons of deviant behaviors.
Extracted, transformed, and cleaned data for four projects using Hadoop HDFS, SQL and Python to support analytics delivery.
Provided weekly supportive data and KPIs to Program Managers to assist project implementation and value capture.
Collaborated with Data Architect team to connect variety of data sources to reconstruct and visualize customer journeys.
03/2016 to 08/2016
Analytics ConsultantNYC Department of Design and Construction － New York, NY
Analyzed bidding data of public construction projects in a team of four to evaluate internal pricing model.
Led weekly meetings with department director to report project progress and present insights using Tableau dashboard and Python Jupyter Notebook.
MS: Business AnalyticsFORDHAM UNIVERSITY, GABELLI SCHOOL OF BUSINESS － New York, NYBusiness Analytics 3.9/4.0 GMAT 720
Center for Digital Transformation Fellowship Award Research Assistant for Business Performance & Risk Management, Big Data Analytics and Database Management
BS: FinanceJINAN UNIVERSITY, INTERNATIONAL BUSINESS SCHOOL University of Wisconsin - Eau ClaireChinaFinance 3.6/4.0 4.0/4.0 Coach of School Basketball Team