FinTech - Cash Application Authority

Backend / REST API / GCP / Firebase / Flask / Pandas / Bash / ETL— Spearheaded transparency creation around data tied to $50 billion in Private Equity (PE) trading potential.— Architected ETL pipelines (GCF/Apps Script/pandas) and DB schema (BigQuery/Firestore) on ~500 PE companies.— Built a mathematical model to statistically describe PE Pricing and market dynamics, which have historically been opaque.— Designed APIs/web backend with Firebase/GCR; collaborated with frontend team to release the first version of product.— Implemented account access control and integrated 3rd-party APIs.

— Collaborated with account managers to perform requirement analysis and technical consulting for Record Management System (RMS)— Day-to-day support and troubleshooting for NJ and PA 150+ police agencies— Provided customized statistic report for clients by MS SQL Server— Enhanced web application reliability/performance and improve database query efficiency

— Identified cross-device anonymous users among 3.3B traffic with 0.6 F1 by Feature Engineering and Random Forest — Enhanced efficiency of ETL pipeline for 1TB data with PySpark and DataBricks— Implemented dashboard backend with FeathersJS and PostgreSQL— Deployed and fine-tuned Apache Druid on Azure Kubernetes Service with 30x speedup for 300M records

— Developed a dynamic pricing AI system using ​ DDPG​ algorithm and ​ PyTorch​ to reduce manpower to 20% and raise gross merchandise volume by 10%— Extracted and cleaned data by efficient ​ SQL​ and ​ pandas ​ from raw transaction data into hourly profiles to train AI— Created backend functions with ​ Python​ on GCP to automatically update prices and databases according to user queries— Conducted customer segmentation research to establish a marketing strategy and raised conversion rate by 6%— Built Google Sheets tools to crawl website data and automate business analysis by Google Apps Script

— Assisted the faculty and students with administrative and technical difficulties and acting as a liaison between the two— Contributed to development of strategies for the program, providing recommendations and deliver weekly presentations— Identified trends through data analysis of course evaluations and cross-functional projects to maximize efficiency

— Crawled, analyzed, and visualized 3000 freedom of information (FOI) requests by Google Apps Script and Python (requests, ​ BeautifulSoup​, matplotlib)— Wrote proposals to raise funds (g0v Grant: 16,000 USD, CIVICUS Grant: 10,000 USD)— Presented in 2019 Taiwan Presidential Hackathon and won Top 5 /132 with the team and suggested better request strategies— Organized workshops to mediate between individuals/organizations making requests and the government officials in charge and improved the process of requesting/publishing government data

— Designed experimental apparatus by SolidWorks and established calibration procedures for air quality sensors— Built the first air cleaner prototype and evaluated its performance; the design was adopted for products— Evaluated performance of air cleaners through analyzing data from monitoring systems by Python