Taiwan
A total 2 years experience of data analysis & data engineer in financial and insurance industry. My project experience including classification ML model training & deployment, database management and data ETL pipeline creation and operation in Hadoop ecosystem. An academic background in Business Administration at National Taiwan University, specializing in the area of operational and business data analysis.
[Data ETL & Warehouse] - Built a full data ETL pipeline to operate over 1100+ tables updated to Hadoop Hive in a weekly basis. - Created a process warning system by both sending messages and providing APIs for the in-company website display. [Data Projects development & deployment] - Developed and deployed 7 projects by PySpark and Linux shell on Hadoop ecosystem. - Maintained 3+ virtual and docker Python environments for Data Analysts running PySpark.
[Anti-Money Laundry(AML) risk prediction model] - Executed a transaction risk prediction ML model for decreasing 65% of human work in low-risk cases in AML. - Rebuilt the model by lightGBM increased 40% the low risk cases accuracy in the same amount of case base. - Created a model monitoring dashboard by Python(Streamlit and Plotly) displayed model accuracy and feature distribution trend and warnings.
- Automated 4 call quality and staff attendance reports decreased 80% of human work by Excel VBA.
[Facebook fan page chatbot] - Built a fan page chatbot by using Botsnova and improved the utilization rate by 70%. - Identified potential customers users by giving tags for their actions. [Facebook digital advertising] - Operated Facebook digital advertising for company fan page posts and videos, and generated 20% more views compared with other interns at the same cost. - Analyzed 2 multinational digital advertising cases help better performance for future advertising for similar situations.