Peter(Wei-Hsiang) Lin Email and Phone Number

• Leveraged AWS services and self-built Python program to build a automation pipeline that sends out signal to Vol planning team, the signal is then used by all stations in the US.• Conducted deep analysis for million-dollar programs to identify improvement opportunities by identifying main features and how they impact the program’s performance under different circumstances with logistic regression and other statistical analyses. (stop-level related projects)• Provided strategic advice on new launched national wide programs by running causal inference analysis. (Central Flow)• Designed and standardized the data science workflow and documentation for AMZL CO BI team.

• Created WBR, MBR, and QBR with Tableau for 400+ delivery stations across North America to track performance and identify top/bottom stations.• Created automation pipeline for next-day forecast and gap the difference between operations plans.• Built Tableau dashboards as one-stop shops for different stakeholders which save 60+ labor hours per week.• Conducted A/B testing (DiD, PSM) to justify new program’s impact on stations across North America, and identify opportunities, which impacted the strategy and main focus of the following 6 months.• Provided deep dive exploratory and statistical analysis to uncover critical factors and identify confounders of a nation-wide program’s performance (about delivery time, delay, on road rescue), and detected anomaly with Light GBM model.

• Acquired US$0.8 million angel funding

Finance:• Determined alternative business solutions with risk evolution analyzed from synthesized financial and budgetary information #Method: Sensitivity Analysis. • Created an analytical framework for monitoring business performance and identifying financial growth opportunities #Method: Horizontal and Vertical Analysis.• Developed annual budgets in collaboration with the development department.-Operation & Business Development:• Clustered domestic companies with web-scraped and government data to help with market segmentation • Conducted independent test on the conversion rate with different industries and service requests from sales leads, and visualized the result to prioritize incoming leads to better target potential clients.• Used decision tree and other predictive models to find decisive factor on sales performance and increased conversion rate from market qualified leads to sales qualified leads by 20%; reduced churn rate by 30%• Led business development team and built solid connections between 30+ clients generating 80% of the company revenue via strategically join targeted exhibitions• Collaborated with the engineering department to improve the project management process. • Successfully earned and conducted a government-supported project with about $35,500 subsidy.

Client: Fortune 500 company• Created a multiclassifier to identify the likelihood of potential customers becoming different types of customers to help our client's sales representatives better prioritize reaching out to new leads.• Created a multiclassifier to identify the major product categories current customers are going to purchase to help our client promote the right product to the right customers with 0.93 AUC (stacking with lightGBM, random Forest, XGBoost).---Client: Hennepin County • Designed two matrices to measure performance of government assistance program in Hennepin County and identify tracts with urgent needs to help client reallocate resources more effectively• Web-scraped online data using Python to establish assumptions that help create analytic framework• Segmented and visualized tracks performance across the County based on two metrices • Used ANOVA, logistic regression, and other models to identify the decisive features and predict the performance for each tract---Client: Famous Coffee Shop• Conducted customer segmentation by clustering and exploratory data analysis, and identified loyal customers and their contribution to the total revenue• Used association rule to create bundle sale package and discount strategies to increase revenue and smooth the demand across different time, day, and month---Client: Customer Experience Intelligence Company• Designed three matrices with psychological support to measure agent’s performance from both transcript and audio data of customer calls• Created agent report card to quantify relative performance of each agent, and visualized the performance of all agents in one call center

ZHEN PIN Co., Ltd: Mobile game startup—• Refined report dashboard and reduced unnecessary buttons by 40% to meet product team’s need via communication with stakeholders and RD team• Conducted A/B testing with time-series analysis to evaluate the impact from new-added feature, and gave suggestions to achieve a 200% increase of DAU in two months• Developed business opportunities by market and customer analysis via quantity and quality surveys and cross-industry alliances. • Designed revenue model corresponding with the business model.

Leverage Inc: Online education platform —• Launched 5 products by gained analytical insights from documented customer responses to key features and figured improved product-market fit via customer-focus survey• Surveyed customers with factor analysis to better understand customer behaviors and advised new official website design• Prioritized a roadmap in order to achieve product goals and metrics.

• Developed marketing plans based on market research and analysis.• Directed and coordinated marketing activities to promote products and services.• Worked closely with all product development departments to create marketing materials.