Brad Stur, an assistant professor at the University of Illinois Chicago, discusses his work on using mathematical optimization to improve the security and public confidence in vote counting. His research, in collaboration with Alex Halderman and Bron Krims, has led to the development of an algorithm that optimizes logic and accuracy testing for voting machines. This testing is crucial for detecting misconfigurations that could lead to incorrect vote counts. The algorithm uses mixed integer linear programming to create test decks that are guaranteed to detect any misconfiguration that swaps votes between candidates, both within and across contests. This approach has been piloted in Michigan elections and is being considered for statewide deployment. The algorithm is designed to be practical, minimizing the number of ballots needed for testing while ensuring comprehensive detection of potential errors. The presentation also highlights the importance of deterministic guarantees in election security and the potential for further research into broader threat models and randomization techniques.