Minjie Wei

Software Engineer at Moveworks

Greater Seattle Area

About

Experience

  • Software Engineer at Amazon Web Services
    Feb 2017 - Present · 9 yrs 6 mos

  • Database Development Intern at Cardlytics, Inc.
    May 2015 - Aug 2015 · 4 mos

    ●Worked backend for several go production projects, designed part of the database and implemented different data engineering process to the projects by SQL. ●Designed and implemented algorithms to merchant map reduction, which reduces query time by 40%.

  • Research Assistant at Tsinghua University
    Nov 2013 - Jun 2014 · 8 mos

    Learning Based Customer Behaviors Analysis in B2C E-commerce Logistics ● Analyzed customers’ preference on various levels combination of logistics attributes like shipping time, shipping rates, delivery method and brand of express company on B2C E-commerce logistics data. ● Applied the choice based conjoint (CBC) analysis method with Hierarchical Bayesian (HB) model in the survey and data analysis process. Implemented Metropolis-Hastings algorithm for model parameter estimation and gave quantitative analysis results for different shipping attributes and levels.

  • Engineering Intern, Design & Industrial Department at Schneider Electric
    Jul 2013 - Sep 2013 · 3 mos

    Procedure Optimization for Label Machines ● Worked on a project to improve the effectiveness of automatic label machine by optimizing its labeling procedures. ● Analyzed the structure of label machine’s components, formulated a linear programming model and implemented on hardware to find the optimal procedure parameters for the label machine.

  • Tsinghua University ()
    • Research Assistant
      Mar 2013 - Sep 2013 · 7 mos

      Passenger Behaviors Based Intelligent Public Transit System ● Designed agent-based modeling and simulation (ABMS) to represent passenger behaviors. ● Designed and implemented an agent-based artificial urban transit system (AUTS), evaluated the performance of system through passengers' processing time cost by simulation using AnyLogic.

    • Research Assistant
      Sep 2012 - Mar 2013 · 7 mos

      Tag-Aware Recommendation Systems ● Based on two underlying networked-based methods: probability spreading (ProbS) and heat spreading (HeatS). ● Proposed tag-aware methods via adding a tag as a new factor of the recommendation system. ● Led to a more accurate recommender result which satisfies customer’s personal demands.