Maytal Saar-Tsechansky

My research aims to advance AI with the motivation to catalyze AI systems’ positive impact in the world and to mitigate potential harms.

Much of my work now focuses on trustworthy AI to collaborate with and advise human experts, particularly in health care. Our research identifies crucial gaps and develops frameworks that can be relied on by experts to improve their decisions and patient outcomes. 

​

My broad agenda is to advance AI methods to benefit people, society, business and organizational outcomes.  This agenda includes AI frameworks for human-AI collaboration, assessing human’s decision-making performance in domains with limited ground truth data (necessary to monitor whether an AI improves or impedes a human’s decision-making), advancing trustworthy AI, and AI systems that cost-effectively learn from and complement imperfect or biased humans.

​

I am also fortunate to engage in social science research and am especially passionate about reducing partisan hate, fostering democratic values, and advancing initiatives that promote overall well-being.

​

Trustworthy AI Grounded in Practical Contexts

​

My research focuses on developing AI methods that are grounded in real-world contexts to improve decision-making. My co-authors and I have shown how some generic AI methods may not only fail to create value for organizations but can even cause harm in consequential contexts.  We then addressed these gaps by creating frameworks for AI that are aware of the context’s properties and its goals. We've shown their effectiveness in critical areas, including health care, fraud detection, and marketing campaigns. 

My research has been informed and inspired by active collaborations with organizations, businesses, and domain experts. Over the years, I addressed challenges in a wide variety of domains, including health care, the future of work, renewable energy, audit, and finance.

​

My research has been supported by government and industry. I initiated and led the University of Texas at Austin’s Translational AI initiative and am an academic board member of the university’s Machine Learning Lab.

​

I serve as Senior Editor at MISQ and at the INFORMS Journal of Data Science, and as an Associate Editor at Management Science. I am also an Editorial board member of the Machine Learning journal.

​

Ph.D students, postdocs:

If you are interested to work with me, please reach out to me at maytal@mail.utexas.edu

​

Below are descriptions of some of my works. See more details in my CV and Google Scholar profile.

​

In 2014 I co-founded Sweetch, a platform for large-scale prediction, prevention and outcome improvement of chronic diseases.