• Working as the marketplace owner for PBI (Pay by Invoice) largest marketplace – US:– Developed the most fundamental initial credit risk assessment model.– Initiated and led the development of Dynamic Line Sizing (DLS) to proactively adjust the credit line size and make personalized underwriting decisions for existing business customers.– Developed a Predictive Spending Model to predict the future spend of active customers. – Implemented 10+ models into production, experienced in writing effective production code.– Built in-house credit assessment for Spain and Italy Amazon Business Pay-by-Invoice from ground.– Initiated a self-service dashboard to track and monitor customer credit trajectory for WW. – Working closely with business leaders and collaborate with cross-functional teams (tech team, business team and PMs) to provide high-value ad-hoc analytic results. • Training and mentoring Interns:– Mentored a summer intern on an analysis on PBI configuration rate, score the customers’ likelihood of configuring with PBI. Proposed marketing campaign to grant incentives to drive and encourage the configuration and activation.

1. Supporting AV fleet management for capacity planning and network optimization.2. Financial trade-off decision making support. 3. Developing interactive web map application for AVFM.

Working as a data scientist in GDIA (Global Data Insight & Analytics) focus on mobility transportation services. Lead and provide consulting services to Ford’s first mobility business GoRide Health:o Developed a demand forecast tool by using machine learning (CNN) with more than 90% prediction accuracy within business acceptable range that can be used more than 10,000+ times per month.o Delivered market pre-launch analysis, such as: depot locations optimization, service area selection and demand visualization.o Created a booking and cancellation dashboard for operation team by visualizing 100k+ records. o Built interactive tools ranking new customers for business acquisition for 5+ markets by using LASSO regression on more than 20 variables.o Work closely with business leaders and collaborate with cross-functional business units to provide high-value ad-hoc analytic results.

• Constructing non-linear pricing and routing optimization models for shuttle-based ride-sharing.• Utilizing Google direction API collecting the real-world data and practically applied the mathematic model. • Comparing several heuristic algorithms for the large-scale routing optimization on 200+ potential stations.

• Formulated dynamic shuttle pricing optimization algorithm for Chariot (a ride-sharing company in SF).• Generated multinomial logistic regression model to simulate the mode choice process for commuters in NYC.• Collected the public data used for simulating real-world operations and improved the total profit by 50% using the optimal pricing and operating strategies.

• Participated in an O2O (Online to Offline) Group promoting movie tickets through a mobile APP “Life in Hand” (Over 1.5 hundred million downloads). • Applied web crawling using VBA and obtained the location of movie theaters all over China. • Handled the daily movie report by collecting and visualizing the box office and theater information.