Le Su reflects on his time at Stevens as an FE student

After graduating from Stevens six years ago, I still recall the first time I came to Stevens and looked at the shining screens in Hanlon Financial Systems Lab. I told myself that I would do everything I could to become an excellent engineer, analyst, and researcher. Although my domain has shifted from Finance to Telecommunication after I started my career at Verizon, my passion for math and computer science never faded away.

My learning experiences at Stevens changed my understanding of education itself. My original perspective of being a good student was to have a good score on every test and exam, but now I know, besides the near perfect 3.967 GPA, what helped me most in my studies and future career are those problem solving skills, good learning habits, collaboration and presentation skills, and most importantly, the confidence and love for my domains.

My advisor, Dr. Ionut Florescu, once said that “an excellent student should be capable of solving every mystery which comes up during the study, and should be capable of integrating the knowledge using critical yet logical thinking”. That became my motto during my master studies, which also drove me to take all the relevant math and computer science courses to gain a better understanding of the logic behind the theories and how to better implement them in real use cases. With this spirit, I found I can easily have a more in depth understanding of each knowledge point, which made me become the top student in each course. With that confidence and interest, I also created study groups with my peers to help other classmates in their studies, which also helped me deepen my love for the analytics field.

The skills I learned at Stevens played a significant role when I carried them over to my work at Verizon. I joined Verizon as the first data scientist in the device technology department. As a leading company in the traditional telecom industry, Verizon has its legacy ways of operating. By analyzing the pain points of the lengthy manual troubleshooting flow from customer care engineers, I realized that there was a need for breaking the information silos between different service departments, and an automated system for the troubleshooting process. After one year of development, my team successfully delivered a data driven Network User Experience Scoring System which covers troubleshooting from device, network, and customer care for each Verizon customer. This project turned into a flagship collaboration project among those three departments.

Due to the contribution in this project, I became the lead of the data scientist team in the device technology department. During that time, my team published 5 papers and 6 patents in the last three years, and were highly recognized by executives from different departments.

Thanks to the experience I gathered from the PhD research seminar in Stevens, I’m now able to handle each research topic well with my team through brainstorming, literature review, code review, and experiment validation. In addition, the internship experience I had in my last semester in Stevens also equipped me with the full flow of a product’s development, from use case discussion, data exploration, system and algorithm design, all the way to the final product delivery. The skills I learned at Stevens helped me expand my skill set during those projects to cover machine learning, telecommunication, and management all by myself. Without those valuable experiences, I never could have done all of those things so smoothly in my career.

Even though I already graduated from Stevens, the bond between me and Stevens never breaks. My team has attended several career fairs from Stevens and helped several fresh graduates find internships or full-time jobs opportunities at Verizon. I also recently initiated a conversation between Stevens and Verizon to discuss collaboration opportunities through part-time education and collaboration projects.

With the increasing complexity of my work and the trend of AI research in so many industries, I started to get in touch with more and more AI machine learning use cases, including natural language processing in sentiment analysis, AI chat bots, computer vision in auto device testing and sports analytics, and reinforcement learning in smart city traffic control. Although there are a lot of online documents available to learn those concepts and sample cases at a high level, I truly believe a systematic learning of the fundamental math and computer science behind these concepts is the only way to bring my machine learning skills to the next level. After starting my career, I gradually realized that each course I took played a key role in my day-to-day work within the industry.