I am Borun Das, a Doctoral Completion Fellow and Ph.D. candidate in the Center for Advanced Computer Studies at the University of Louisiana at Lafayette. Advised by Dr. Xiali Hei, my research focuses on developing reliable generative AI and machine learning methods for scientific discovery, with particular emphasis on constraint-aware materials design, polymer informatics, and molecular property prediction.
My doctoral research lies at the intersection of generative AI, machine learning, natural language processing, and computational materials science. I develop methods using large language models, conditional generative models, predictive machine learning, and reward-guided reinforcement learning to generate and evaluate scientific candidates under multiple constraints. My recent work investigates how explicit feedback, chemical validity checks, property prediction, and novelty assessment can be incorporated into generative systems, and I am also exploring tool-augmented agentic AI workflows for iterative scientific design and evaluation.
To connect computational research with real materials challenges, I collaborate with Dr. Guoqiang Li at Louisiana State University and Dr. Andrew Peters at Louisiana Tech University. These collaborations allow me to work across computer science and materials engineering, with thermoset shape-memory polymers serving as a primary testbed for developing and evaluating AI-driven design methods.
I previously served as a Graduate Teaching Assistant at the University of Louisiana at Lafayette, supporting courses in Advanced C# Programming and Networking in Java through laboratory instruction, office hours, grading, student mentoring, and selected guest lectures. This experience strengthened my interest in teaching and in helping students connect computational concepts with practical problem-solving.
I hold an M.S. in Computer Science from the University of Louisiana at Lafayette and a B.Sc. in Computer Science and Engineering from Stamford University Bangladesh. More broadly, I am interested in developing AI methods that can transfer across scientific domains, including emerging applications in biomedical materials, medical imaging, medical text analysis, and healthcare decision support. My long-term goal is to build reliable AI systems that combine generation, prediction, evaluation, and human expertise to support scientific and engineering discovery.
My CV: CV
