BY ERIN COGSWELL
Ali Khan (’21) first learned to program from a Liberty Basic for Dummies book when he was just 11 years old. From there, he fed his passion for programming by teaching himself web development through Java — and even how to program his TI-84 calculator.
Fast-forward nearly two decades, and Khan is a second-year Ph.D. student in computer science and engineering at UNT.
He recently earned a $53,000 Graduate Research Fellowship from the National Science Foundation, one of the most competitive research awards in the U.S. for graduate students. The award will fund Khan’s research under his faculty advisor Sanjukta Bhowmick, an associate professor in computer science and engineering, for the duration of his Ph.D. and opens doors for networking with other top-tier researchers and labs across the nation.
“This award is very well deserved,” Bhowmick says. “I have been working with Ali since he was an undergraduate student at UNT, and I am always very impressed with his professionalism, dedication to work and excitement for exploring new areas. Ali also serves as a very effective ambassador for our lab among the current cohort of undergraduate students.”
Khan’s focus is the union of machine learning, high performance computing and graph theory. His research centers on developing scalable neurosymbolic algorithms for network comparison.
Simply put, he tries to understand networks like the internet to help other scientists conduct their research. Khan’s current work looks at quickly troubleshooting problems with the software scientists use on supercomputers. It also can help artificial intelligence be smarter, faster and more energy efficient.
Computer science and engineering always have been a family affair. Khan’s father has worked at Texas Instruments for many years and is an engineer in research and development. Khan also has two cousins who worked at Microsoft. One is still there and is a principal software engineering manager while the other now works at Robinhood as an engineering lead.
Khan picked up an interest in research while his mother, now a provost at Valencia College in Florida, was working on a Ph.D. in English literature. Watching her research process during her graduate study was an early exposure to the work required in higher education. Long before his own doctoral pursuits, Khan put his research skills to use in his high school debate club. While he admits he wasn’t great at arguing on the spot, he excelled at researching and developing arguments to use in the rounds.
“I’m interested in diving deep and understanding how things work,” he says. “I learned to love research while performing undergraduate research. I got involved in research and found the work more interesting and impactful.”
Khan, 28, earned a bachelor’s degree in computer science and engineering at UNT in 2021 before moving into the Ph.D. program. During his time at UNT, Khan has worked on multiple research projects, including RE-PLAN — a cloud-based computer program developed by researchers in UNT’s Center for Computational Epidemiology and Response Analysis to help healthcare officials plan emergency responses for disease outbreaks. During his year on the project, the team pivoted to focus on developing a response for COVID-19. Denton County officials used the system to plan the locations of their vaccination centers, and RE-PLAN is now being commercialized by Juvare, a leader in critical incident preparedness and response technology.
For his undergraduate thesis, Khan combined traditional graph algorithms with advanced deep learning models for natural language processing — the technology that powers tools like ChatGPT. Khan and his team developed a multilingual tool to help researchers summarize and understand long scientific articles and books.
“My professors, including Dr. Bhowmick, Dr. Marty O’Neill and Dr. Paul Tarau, helped me realize that I can have flexibility in focusing in the areas that I find interesting while still making a meaningful impact to the rest of the world,” he says. “I’m fortunate to be working with these professors, who are leaders in their respective fields. It’s been a positive experience that has exposed me to many new ideas and research opportunities.”
At the 2022 SuperComputing Conference in Dallas, Khan met one of Bhowmick’s collaborators, who encouraged him to apply for an internship at Sandia National Laboratories in Albuquerque, New Mexico. This summer, Khan was there focusing on emerging hardware accelerators from Google, GraphCore and others that speed up AI training. Specifically, he studied sparse tensors, which enable efficient data storage and processing, to ensure they’re performing optimally for this new technology.
“My experience at Sandia was positive,” he says. “I was able to interact with both my peers and leaders in the computational fields. Sandia is a strong research institution, so I enjoyed being exposed to what others are focusing on.”
With multiple internships under his belt, Khan encourages undergrads to be involved on campus as much as possible.
“Attend hackathons, coding competitions and UNT’s AI Summer Research Program,” he says. “Go to as many summer internships as you can and assist your professors with their research. Even if you don’t go into academia, these experiences will be invaluable to your future endeavors.”