Image
Sheng Li and Chirag Agarwal Named 2025–2027 Capital One Fellowship Awardees
The University of Virginia School of Data Science is proud to announce that Professors Sheng Li and Chirag Agarwal have been selected as recipients of the 2025–2027 Capital One Fellowship Awards, which provide faculty with two years of support for an outstanding PhD student, named a Capital One Fellow. This partnership between the School of Data Science and Capital One strengthens the school’s research in artificial intelligence and machine learning while preparing future data scientists to address complex challenges in finance and beyond.
Sheng Li, Quantitative Foundation Associate Professor, leads the Reasoning and Knowledge Discovery (RISE) Laboratory, where he focuses on developing intelligent systems for open and dynamic environments. His research spans trustworthy machine learning, generative AI, computer vision, and causal inference. His fellowship project, Integrative Decoding for Reliable Large Language Model Reasoning in Financial Services, aims to improve the factual accuracy and efficiency of large language models in high-stakes financial tasks such as credit analysis and fraud detection.
Mengxuan Hu, Li’s Ph.D. in Data Science student and Capital One Fellow, will build on her prior research in safe and aligned AI to design reinforcement-learning–based decoding policies tailored for financial tabular reasoning. Her work explores how decoding-time interventions, guided by domain-specific reward functions, can enhance factual accuracy and decision quality in real-world applications.
Image
Chirag Agarwal, Assistant Professor and director of the Aikyam Lab, studies trustworthy machine learning frameworks that advance explainability, fairness, and robustness in AI systems. His fellowship project, Towards Multimodal Graph-Language Reasoning, introduces new benchmarks and architectures to help graph-language models better integrate structured and textual data for tasks such as fraud detection and risk prediction.
Ding Zhang, Agarwal’s PhD student and Capital One Fellow, will focus on explainable and trustworthy AI for modeling complex graph data. His research aims to make AI systems both interpretable and reliable in sensitive domains like healthcare, finance, and public policy.
The Capital One Fellowship Program reflects a strong and growing partnership between industry and academia, advancing research at the intersection of artificial intelligence, data science, and financial technology.
“At the School of Data Science, we believe the future of higher education depends on trustworthy relationships with the private sector,” said Phil Bourne, founding dean of the UVA School of Data Science. “Our long-standing relationship with Capital One exemplifies that belief, and this latest partnership with two of our outstanding faculty speaks to its promise.”
