People
  /  
Students
  /  
Class of 2023

Photo of Zhengyuan (Donald)  Li

Zhengyuan (Donald) Li

Graduate StudentEmail Zhengyuan (Donald) Li

Zhengyuan (Donald) Li graduated in 2018 from Emory University with distinction. At Emory, he completed a double major in business and computer science and developed a passion for solving real-world business problems with technology. Donald was first exposed to the concepts of analytics while working for EY’s decision modeling team as summer staff. He taught himself Power BI to transform massive data from complex scenario analyses into clear and telling visuals for an elevator pitch to a tourism giant’s c-suite. This experience kindled his interest in generating data-driven insights for companies in the T&T space. Thus, Donald joined Carnival Cruise Line with its Revenue Management Science team as a business data analyst/decision scientist after graduation. During his four years working for Carnival, Donald acted as an internal consultant for the product team to maximize revenue using Python, SQL, and Tableau. Through daily communication with business stakeholders, he learned that communicating technical details in a fashion that a non-technical audience can easily comprehend is just as important as the engineering know-how itself. Upon this realization, he took the initiative to build an interactive dashboard that translated existing heuristic thinking into an interpretable CART algorithm, saving overall decision-making time by 90%. Inspired by the success of this end-to-end pricing analytics solution, the management team decided to automate revenue management for the company’s premium lines of products, resulting in a $16 million revenue increase YoY. Understanding that analytics techniques would help pump fresh blood into the vessels of traditional industries, Donald educated himself on the fundamentals of machine learning. He spearheaded a team of student consultants to increase promotion targeting accuracy. The team replaced the original blanket email campaign with an automated EDM workflow based on a voting classifier that combined logistic regression, random forest, and XGBoost, thereby achieving a 2% engagement rate uplift while reducing the total number of emails sent by 30%. Through Northwestern’s MSiA* program, Donald aims to further build upon his theoretical knowledge and application toolkit. For him, real-world analytics solutions extend beyond models - from understanding the business domain to proposing cutting-edge and, most importantly, relevant techniques, and finally, to building an elegant implementation with low communication costs, an analytics project requires data science practitioners to leverage both hard skills and soft skills. Donald is looking forward to collaborating with industry professionals and exploring different possibilities for his analytics career.

*later renamed MS in Machine Learning and Data Science