As an undergraduate, Sahinidis studied chemical engineering at Greece’s Aristotle University where he first realized that he particularly enjoyed math and writing code. Through these interests, he discovered optimization and its many applications. Then, when he was in graduate school – earning a doctorate in chemical engineering – at Carnegie Mellon, Sahinidis encountered the academic journals Operations Research and Management Science, and he read every past issue he could get his hands on. He also took business classes in mathematical programming, which led him to integer and linear programming. 

“I tried out optimization, and I loved it, so I stuck with it. By the end of my graduate studies, I was firmly entrenched in systems engineering, and my first faculty appointment was at the University of Illinois Urbana-Champaign in the industrial engineering program, where I taught basic operations research classes,” Sahinidis reflected. 

He had learned that optimization was helpful for efficiently planning and scheduling chemical processes – and much more: supply chain management, airline scheduling, and device designs. He explained, “The math underneath is the same. Application agnostic algorithms can be applied by many people in different domains, and that was what particularly fascinated me.” 

Sahinidis spent 13 years as a professor at his alma mater as the John E. Swearingen Professor of Chemical Engineering. During his tenure at Carnegie Mellon (where he still holds a courtesy appointment), he continued the development of BARON (Branch-and-Reduce Optimization Navigator), a global optimization software system. BARON solves challenging nonconvex optimization problems, including continuous, integer, and mixed-integer nonlinear problems. Sahinidis also created ALAMO (Automated Learning of Algebraic Models), which is a black-box modeling tool that generates simple yet accurate algebraic models from data. 

Given his unstinting interest in optimization, it is perhaps unsurprising that Sahinidis would eventually arrive at the No. 1-ranked Stewart School, with its renowned optimization researchers. 

“ISyE has the world’s best optimization group – and a top machine learning group, too. I am excited that I am working with [Institute Emeritus Professor] George Nemhauser and [A. Russell Chandler III Professor] Santanu Dey on some problems – linear, mixed, and nonlinear – that our field has struggled for decades to solve,” said Sahinidis. “There’s a famous collection of challenging test problems in this area that originate from applications, including nuclear reactor management, facility location, pipeline design, and other engineering problems. When I first started working on them, we were able to solve about 5-10% of them. Now, we can solve about two-thirds of those problems, and I’m hoping that what we’re researching with Santanu and George will push the capabilities of optimization solvers for these problems close to 100%.” 

Specifically, Sahinidis’ current research activities are at the intersection between computer science and operations research, with applications in various engineering and scientific areas, including theory, algorithms, and software; informatics problems in chemistry and biology; and process and energy systems engineering. 

The ability to perform significant interdisciplinary work with other Georgia Tech faculty members also drew Sahinidis to ISyE, and his affiliate appointments at the Institute underscore this: He is also involved with the Algorithms, Combinatorics, and Optimization Program; the Institute for Data Engineering and Science; the Parker H. Petit Institute for Bioengineering and Bioscience; the Manufacturing Institute; and the Strategic Energy Institute.  

“Optimization is ubiquitous in applications in science and engineering. With the recent advances in machine learning, optimization is fueling developments in areas we never imagined we could address before. It is truly exciting to be at Georgia Tech, where I can collaborate with world-class colleagues and graduate students on optimization and its applications,” Sahinidis concluded.