Python, Machine learning, Data Science, C++, Java

Jianxiang (Gary) Gong

929-***-**** ****.******@*****.*** linkedin.com/in/jxgong github.com/garyjxgong EDUCATION

Fordham University New York, NY

MS, Quantitative Finance (STEM) GPA: 3.8/4.0 08/2019 - 05/2021

• Concentrations: Machine Learning, Quantitative Trading, Python, C++ University of Washington Seattle, WA

BS, Mathematics 09/2014 - 08/2018

• Concentrations: Statistics, Probability, Linear Algebra, Data Structures and Algorithms SKILLS

• Programming: Python, SQL, R, Java, C++, SAS

• Frameworks: TensorFlow, Scikit-Learn, PySpark, Scrapy, REST

• Tools: Spark, SageMaker, Anaconda, Jupytor, MySQL, PostgreSQL, MongoDB, Git, Linux, Tableau

• Modeling: Regression, Tree-based models, Neural Networks, Time-series forecasting EXPERIENCES

WhatsBusy, Data Scientist Intern 05/2020 - 08/2020 A platform helps 100K+ restaurant owners to maximize profits

• Analyzed POS data using SQL and Python and communicated findings and recommendations with stakeholders

• Built a 0.91 F1-score Random Forest model in Python to uncovered root causes of low-performance employee

• Grew revenue for pilot users by 3% MoM by integrating an item recommender to increase employee sales

• Accelerated data aggregation by 50% with Python list comprehension to avoid repeatedly calling append attributes

• Reduced data defect to 2% via verifying with 700K US restaurant data crawled using Python(Scrapy) PROJECTS

Cryptocurrency Trading System (Python, SQLite, Websocket) 07/2020-12/2020 An end-to-end trading system that allows both backtesting and live trading

• Designed ETL process with Python and SQLite to receive, sample and persist data from broker's APIs

• Optimized the base data loader in Backtrader library using Python to allow machine learning trading signals

• Developed a Python script to stream real-time tick data from broker using Websocket to facilitate live trading

• Established a 1-D Inception Net in TensorFlow framework trained on 440K+ time-series patterns to forecast trend switching point of Bitcoin with 61% precision

Self-Driving Car (Python, Tensorflow, OpenCV) 09/2020 - 11/2020 An autonomous vehicle AI trained in the game Grand Theft Auto V

• Constructed a data pipeline in Python to perform real-time lane detection with OpenCV during the game

• Trained Nvidia's DAVE-2 network on 30 minutes hand-collected video to mimic driving behavior and adopted pre- trained YOLOv3 for object detection using TensorFlow

• Improved training memory efficiency by customizing a Python Generator to process batched data on the fly Mortgage Default Detection (Python, Scikit-Learn, Pandas) 03/2020-05/2020 A machine learning model to identify high-risk mortgage applicants

• Investigated 2.6GB of mortgage data using Python(Pandas) and conducted visualization using Seaborn

• Created ten new features by domain knowledge feature engineering, which boosted base model ROC by 5%

• Improved the ROC from 0.75 to 0.78 in identifying high default risk applicants by stacking Random Forest, XGBoost, and LightGBM model after tuning with Grid-Search and Cross-Validation

Contact this candidate