Jae Huang

Experienced data engineer with Python, database and data pipeline.

Taipei，TW | New York，US

Skills

Python / R / Java

SQL: MySQL / SQLite / Postgresql

GUI: TKinter / PyQt

Data Cleaning / Data Standardizing / Feature Engineer / Data Modeling / LargeData Handling / Data Visualization

macOS / Linux / Windows

BI Tool: Tableau / Google Data Studio

Self-learning: AWS / Docker / Airflow / MongoDB / Hadoop

Optimized applications on operation automating process, reduced data processing time 50-70% for each project
Designed operation dashboard to keep track on data collecting progress, built a Python-based data pipeline to keep updating new coming data and data integrity
Supported on consuming behavior analysis project with large data processing and user predictive analysis; designing an application to auto-generate statistic report for ad effect
Standardized the quarterly-recurring sampling work by designing an sampling application, simplified 50% of the steps

Built marketing dashboard with Advanced Excel, SQL and Google Analytics for a weekly average of 80 new patients from various sources Electronic health record system, Google Business Insight, Zocdoc etc.
Leveraged R to design performance report to measure patient retention rate and improved from 4 to 8 visits per patient in 2018
Optimized patient review on 4.5+ rating by discovering provider's strength and improving weakness on Excel analysis
Delivered patient volume report with data visualization to stack holder to understand business efficiently, discovered seasonal pattern that increased 20% of new patient

Developed course-job recommendation model for APD training company by helping users choose courses to match their career goal, and calculate the likelihood of getting a job based on users' educational background
Used Selenium to crawl information from NYU CareerNet and NYU SPS school websites, reduced 90% processing time
Processed 2, 000 jobs and 446 courses text files through Python for data preparation and prototypes design
Quantified the similarity between course and job using different algorithm models tf-idf, LSI, etc. improved prediction accuracy from 30% to 67%.
Determined the model to apply on the defining business problem by model selection technique KL Divergence
Built and evaluated models to predict the probability of getting a job based on users' educational background

Initiated KPI tracking system - Designed an easy to use dashboard to track the KPI achievement progress along with budget control; conducted an achievable timetable to reach the year goal.
Implemented web analytics - Identified the demographic of the kindness community; optimized the retention on website by recognizing user behavior with Google Analytics.
Constructed social media analytics - Using Python to build automatic generated social media report reducing 70% time on report; recognized the most effective Facebook fan page promotion; analyzed the cost and ad for fan page to support constructing the 2017 fall community growth plan.

Detected irregular web traffic and identified the leading cause by using Yahoo web analytics; kept sales run at 9M weekly
Developed Facebook fan page growth strategy through social media marketing analysis; led to page likes increased 200, 000 in 5 months, engagement grew 476%, monthly sales grew 310%

GPA: 4.0

Client Project – Ports America Empty Container Analysis

Provide business insight into the company operating environment with Tableau and SQL.
Identify leading causes of the company resource wasting problem by using R and statistics tests.
Used machine learning model (e.g. random forest, regression, neural net work and SVM) to perform prediction.

Graduation Project