James Zhibin Sun

Data Scientist, Software Developer at Pegasus Vertex, Inc. - Drilling Software Company
  • Claim this Profile
Contact Information
us****@****om
(386) 825-5501
Location
Houston, Texas, United States, US
Languages
  • English Native or bilingual proficiency
  • Chinese Native or bilingual proficiency
Skills

Topline Score

Topline score feature will be out soon.

Bio

Generated by
Topline AI

You need to have a working account to view this content.
Join now
You need to have a working account to view this content.
Join now

Credentials

  • Google Analytics Individual Qualification
    Google
    Feb, 2018
    - Nov, 2024
  • Google Adwords Search Certification
    Google
    Mar, 2018
    - Nov, 2024
  • Associate Financial Planner
    Certified Financial Planner

Experience

  • Pegasus Vertex, Inc. - Drilling Software Company
    • United States
    • Oil and Gas
    • 1 - 100 Employee
    • Data Scientist, Software Developer
      • Jan 2019 - Present

      • Predictive Modeling: Trained a regression model using Random forest on acquired drilling data from the field to predict drilling rate of penetration, helping engineers adjust drilling parameters to maximize drilling speed • Predictive Modeling: Trained a regression model using XGBoost to predict displacement efficiency in drilling well cementing, researched and used SHAP values (SHapley Additive exPlanations) to interpret the model by calculating and visualizing feature importance • Chatbot: Researched Rasa, a Python-based AI framework, and developed a chatbot for company’s new website to provide automated text conversations purely based on text input. Deployed on a CentOS Linux server using Docker and implemented version control using GitHub • Natural Language Processing: Applied NLP methods to analyze drilling mud text data, including stop words removing and word stemming. Developed TF-IDF analysis in company’s MudManager and MUDPRO+ 3.0 software to find incident related keywords. Discovered information buried in the text data that was not exploited in the industry before • Natural Language Processing: Researched and developed collocate word cloud, word hierarchical clustering, and word association analysis and visualization in the latest MUDPRO+ 4.0 software using NLP methods • Developed inventory prediction and alert module in MUDPRO+ 4.0 software using time series analysis • Solved the problem of integrating R scripts into .NET Framework using R.NET without requiring customers to install R • Developed modules in other software, including UI, calculations, plots, and report documents generation using C#, VB.NET and JavaScript • Designed software requirement with product managers and software source code deployment and delivery • Researched applications of AI in drilling, summarized and gave presentations to CEO and PhDs in petroleum Show less

  • Wizard Wall®
    • United States
    • Retail Office Equipment
    • 1 - 100 Employee
    • Digital Marketing Analyst
      • May 2018 - Aug 2018

      • Developed a web crawler using Selenium in Python to automatically calculate shipping fees on UPS website for precise cost estimation • Initiated website digital marketing campaigns, improved performance of Google and Facebook ads by A/B test and visitor data analysis using Google Analytics, increased website sales threefold • Developed a web crawler using Selenium in Python to automatically calculate shipping fees on UPS website for precise cost estimation • Initiated website digital marketing campaigns, improved performance of Google and Facebook ads by A/B test and visitor data analysis using Google Analytics, increased website sales threefold

  • Bank of Shanghai
    • Banking
    • 200 - 300 Employee
    • Financial Planner
      • Jul 2011 - Mar 2016

      • Managed private wealth management accounts, analyzed CRM system data and transaction patterns to identify customer segments and high risk and low risk profiles, matched customers with suitable product portfolios • Compiled quick reference guide and led training programs for new banking system in 9 branches, achieved shortest system conversion time • Developed mobile social media public accounts as alternative marketing channel, engaged staff from other departments in webpage sharing, reduced the time delivering information to clients from one week to one day Show less

Education

  • University of Maryland College Park
    Master of Science, Information Systems
    2017 - 2019
  • Donghua University
    Bachelor of Science, Finance
    2007 - 2011

Community

You need to have a working account to view this content. Click here to join now