Hanoi student shifts from chemistry to AI, wins $370,000 US scholarship in computer science

Nguyen Manh Duong, 22, recently completed his data science and artificial intelligence program at Hanoi University of Science and Technology (HUST) with a cumulative GPA of 3.94/4, graduating with highest honors. Each year, only about 7–8% of HUST students graduate with distinction (3.6/4 or higher), and very few surpass 3.9 as Duong did.

Before officially receiving his bachelor's diploma, Duong received, in March, an admission letter to the PhD program in computer science at the University of Illinois Urbana–Champaign, along with a full scholarship package worth about US$370,000 over five years. According to U.S. News & World Report, the university ranks 33rd overall in the U.S. and 5th for computer science.

From a chemistry-specialized student at Hanoi National University of Education High School for the Gifted to a HUST undergraduate, Duong had never planned to study abroad or pursue a research career. "It all came like destiny," he said.

Duong entered university in 2021, when artificial intelligence was still relatively nascent in Vietnam. After researching the field, he saw both promise and purpose, and decided to switch.

He had gained admission to HUST via his high school academic profile and interview, bolstered by a second prize in the national science and engineering competition in chemistry. Surrounded by peers with stronger foundations in math and computer science, he initially felt insecure—but that became motivation to master AI as deeply as possible.

Nguyen Manh Duong in a photo he provides.

"I never studied just for grades. I always thought I had to study to really understand," Duong said. He constantly asked questions and sought answers on his own, which helped him grasp concepts thoroughly and broaden his knowledge. That approach earned him mostly A and A+ grades, even in demanding courses like calculus and statistics, keeping his GPA above 3.8 from his first year.

Beyond academics, Duong was active in extracurriculars, serving as deputy head of the school's events and programs club. In his second year, he entered the "Mr. & Miss HUST" contest and finished as runner-up. He also interned as an AI engineer, but found the role limiting: there was little room for exploration.

"Even though my grades were among the top in my class, I felt lost. Working for a company did not suit me, but I did not think I was capable enough for research either," he recalled.

In his third year, he reached out to HUST's AIoT Lab to try research. Applications had closed, but he emailed Associate Professor Nguyen Phi Le, then head of the lab and now head of the AI4LIFE Research Institute, and, after an in-person interview, was accepted. Research proved more challenging than coursework; it demanded constant reading as AI evolves daily.

After three to four months, with guidance from Dr. Le and senior lab members, he adapted. "She taught me patiently—from how to read a paper to knowledge that would have taken me years to figure out on my own," he said. At the lab, he also collaborated with various professors and experts who helped him refine his direction.

Focusing on federated learning and medical AI, Duong became lead or co-author on seven publications at reputable international venues, including one at the International Conference on Computer Vision, one of the world's top computer vision conferences. By the end of his third year, he began considering graduate study abroad.

Aiming to strengthen his research record, he initially planned to apply after graduation but this year submitted a single application to the University of Illinois Urbana–Champaign, based on two criteria: a top-five U.S. computer science program and a lab specializing in medical AI. During a joint HUST–VinUni project, he had also met several Illinois professors.

He spent about three weeks preparing his application: transcripts, research portfolio, recommendation letters, and a personal statement outlining his motivations and goals. Three professors in Vietnam and the U.S. who had mentored him wrote recommendations. "When I got the acceptance, I was overwhelmed. I honestly was not very confident, since I started research much later than many others," he said.

Dr. Le said Duong is one of three AIoT Lab students this year to win a fully funded U.S. PhD in computer science before graduating. "He is an all-rounded student, with logical thinking, research ability, and soft skills," she said.

Duong flew to the U.S. last month, about a month before receiving his bachelor's diploma.

For now, he continues collaborating with the AIoT Lab while equipping himself with socio-economic knowledge for broader perspectives abroad. "I hope to publish collaborative research with HUST labs while studying in the U.S.," Duong said.