Kai Wang 2023 Early Career Fellow
Affiliation Assistant Professor, Georgia Institute of Technology Hard Problem Made game-changing contributions by having AI address one or more of humanity’s greatest challenges and opportunities.

Kai Wang received his Ph.D. in Computer Science at Harvard University working with Professor Milind Tambe, and Kai will be joining Georgia Institute of Technology as an assistant professor in the School of Computational Science and Engineering in Spring 2024. His research interests include AI for social impact, multi-agent systems, computational game theory, machine learning, and optimization, with a focus on challenges in health and environmental sustainability. One of Kai’s key contributions is the study and real-world deployment of decision-focused learning to provide intervention recommendations for societal challenges. Kai proposed integrating decision-making processes learned from stakeholders into machine learning to leverage domain knowledge and improve learning performance. Specifically, in collaboration with an Indian non-governmental organization, ARMMAN, Kai designed the first decision-focused learning algorithm for a mobile maternal health program with 2.9 million users to improve awareness of health information and reduce maternal mortality. The algorithm has been deployed and is currently used by more than 100,000 people with a 31% improvement in user engagement. In recognition of his technical contributions and real-world impact, Kai is recognized with the Siebel Scholars award and the best paper runner-up award at AAAI.

AI2050 Project

Kai Wang’s AI2050 project centers on leveraging decision-focused AI to address societal challenges, particularly in health and environmental sustainability. The decision-focused AI approach integrates machine learning with optimization to train models based on decision quality, leveraging domain knowledge from decision-making processes in high- stakes domains to improve model performance. The goal is to strengthen decision-focused AI to conquer more societal challenges and collaborate with stakeholders to deploy AI solutions for social impact.

Project Artifacts

G. Kornowski, S. Padmanabhan, K. Wang, Z. Zhang, S. Sra. First-Order Methods for Linearly Constrained Bilevel Optimization. arXiv. 2024.