• Written SQL scripts of launching new tables for the specific customer needs analysis or optimized existed scripts for adding customer portraits labels(columns) into Vertica Database and capsuled SQL scripts into Perl scripts for DBA to test and update on ETL platform• Written prepositional and postpositional dependency for launched scripts, and updated new labels to Big Data Labels Management Platform and finished the data audit• Updated daily/weekly/monthly data to the Vertica database through DbVisualizer client program via SQL• Written SQL scripts and capsuled into Perl scripts to fetch different themes data (ex:5G) for specific analysis• Used Impala component, and written SQL and HiveSQL scripts to fetch data from Hive and Vertica databases to build models for targeting, suspending service and analyzing telecommunication fraud users

Teaching Assistant of Marketing Analytics

Teaching Assistant of Big Data I & II

• Collected and cleaned the finance income and spending data of 2000 regions and 300 cities in China, analyzed it exploratorily, calculated and prepared the potential variables for modelling • Selected and normalized the variables from the dataset, built and optimized different clustering models in R using algorithms including k-means, DBSCAN and hierarchical clustering• Determined the best algorithm for a score which measures each region and city’s comprehensive economic power, and used the score as the reference to select the best model• Found the best cluster model in which the different regions and cities were assigned into clusters with different financial risk levels, the model helped the company to decide whether to invest projects in certain areas and might help to save millions of Yuan

• Analyzed operation data, built charts and wrote the operation analysis report monthly, presented the results and conclusions of the operation analysis in the department meeting• Analyzed the data and built the model for new rules of bankers’ performance assessment and salary rating• Collected and analyzed operation data of local branches and competitors, wrote research reports