The Schedule Tamer

 Picture this familiar nightmare: A group text with friends about “grabbing dinner” quickly spirals into a mathematical impossibility, with five conflicting schedules, three restaurant vetoes, and a quadratic equation of dietary restrictions. Now imagine a similar scheduling challenge with 30 professional sports teams, thousands of games, and millions of dollars in media contracts at stake. Welcome to the world of sports scheduling.

“It’s one of the most difficult mathematical problems there is,” says John Stewart ’97, serial entrepreneur and co-owner of Fastbreak AI. Over the past several years, Stewart’s AI-powered company has cracked the code on sports scheduling and become, in his own words, the “800-pound gorilla” in the industry. Beyond solving a complex mathematical puzzle, Fastbreak AI’s technology is transforming how millions of fans experience sports—from ensuring NBA showcase games appear in prime-time slots to improving the experience for young athletes playing in daylong tournaments. But the path from WPI mechanical engineering student to sports tech innovator wasn’t a straight line, and Stewart’s story is one of continual reinvention, driven by an engineering mindset and an entrepreneur’s eye for opportunity.

For nearly 100 years, the task of scheduling games for professional sports was done largely with pencil and paper, including by a husband-and-wife team who scheduled games for Major League Baseball out of a home office. Henry and Holly Stephenson spent 25 years scheduling 2,430 annual games played by 30 teams—and though they began their planning with a custom-built computer program, details were ironed out by hand. “[N]o matter how simple a schedule format looks, they all end up terribly complicated,” Henry once admitted to Sports Illustrated. Over time, those complexities grew, with interleague play and media rights adding considerations beyond what the couple could manage.

In 2004, the MLB hired a sports scheduling company to add a layer of polish the Stephensons couldn’t achieve. The NBA, which had also employed the Stephensons in the late 1970s, brought their scheduling in-house until deciding in 2015 that the task had outgrown their capabilities and signed a contract with accounting juggernaut KPMG. There the task was led by Chris Groer, a mathematician with expertise in optimization problems.

In July 2022, Stewart, Groer, and Carnes launched Fastbreak AI, bringing cutting-edge AI technology to one of sports’ most challenging behind-the-scenes operations. The National Hockey League became a client in October 2023, with the NBA coming on board just five months later. Today, 17 of the top 20 professional leagues are clients, in addition to major collegiate conferences like the Southeastern Conference and the Big East. Revolutionizing the $40 billion youth sports market is also on the horizon.

Fastbreak AI’s innovation lies in teaching algorithms to recognize what humans instinctively feel makes a “good” schedule. “If you’re a league scheduler with plenty of tribal knowledge and I ask you if this is a good road trip, you might say, ‘Yes, it’s a good road trip—but I don’t know why. It just is,” Stewart explains. “The problem is you can’t really program ‘I don’t know, it just is’ into a computer.” Fastbreak AI’s solution involves having league experts rate schedules on a scale of 1 to 10, while the system analyzes the many variables behind each schedule. This process gradually teaches the model what constitutes quality. “We’ve gotten so good that we can predict the quality score of the road trip better than the humans who programmed it,” Stewart notes.

While Fastbreak AI launched just three years ago, its foundations were many years in the making—a convergence of Stewart’s engineering background, Groer’s mathematical expertise, and Carnes’s optimization skills. But to understand how these three innovators arrived at this intersection of sports, technology, and problem-solving, we need to look back to where Stewart’s journey began.

Decades before launching Fastbreak AI, Stewart was a young man captivated by the mechanics of transformation. “My favorite toys growing up were Transformers,” he recalls. These shape-shifting robots sparked a fascination with how things work and change form. His early interest in robotics revealed an aptitude that would later define Stewart’s career—finding the underlying systems that make complex things work. The natural affinity he showed as a teenager led him to mechanical engineering, a foundation for seeing opportunities where others saw only problems.

His career choice invited professional insights from his father, an architect, as well as some good-natured ribbing. “He liked to tell me that an engineer is just an architect without a soul,” he recalls with a chuckle. Following high school, Stewart chose WPI to begin his training as an engineer—not just for its rigorous academics, but also its focus on concrete applications. “My professors were all industry experts and they tied what we were learning to the real world,” he explains.

This real-world orientation culminated in Stewart’s Major Qualifying Project, where he worked on-site at Pratt & Whitney, the Connecticut-based aerospace manufacturer, developing a fixture design for a protective coating system. Stewart leveraged this experience to establish himself in the industry while still a student on the Worcester campus.

“I worked for a bunch of engineering companies while I was still at WPI, and I got heavily into finite element analysis (FEA) at MSC Software.” At its core, FEA works by breaking down a complex object into smaller, simpler parts called “finite elements.” Instead of trying to solve one complex equation for the entire structure, FEA solves simpler equations for each small element and then combines these solutions. Stewart’s experience with FEA foreshadows the approach he would eventually bring to the complexities of sports scheduling.

From his WPI education to his hands-on engineering experience, Stewart developed a mindset that would guide his entire career—not just in solving technical challenges, but in building companies ready to scale. “Engineering is really about continuous learning, and particularly with technology, if you’re not always in a learner’s mindset, you’re going to be left behind really fast,” he says. “The project-based approach I learned at WPI is absolutely applicable to how we run our company day to day.”

This approach would prove invaluable as Stewart began his entrepreneurial journey, launching his first company and discovering his talent for identifying market needs and assembling world-class teams to meet them.

Less than 12 years after his graduation from WPI, Stewart prepared to sell his first company, Saber Design and Analysis, a 60-employee engineering firm that took advantage of his FEA expertise. Before the ink had dried, he began plotting his next venture, a consulting practice built on Salesforce—a cloud-based customer relationship management platform that had become essential for businesses of all sizes. In his new business, Stewart fielded a simple request from a client in construction waste management: to see where the company’s dumpsters were on a map. This seemingly modest need sparked the idea that would evolve into MapAnything, a software application that eventually transformed how businesses visualize and optimize their geographic data.

The concept for his initial product was rudimentary—“barely a Google Maps integration,” Stewart says—but once uploaded to Salesforce’s app store, it revealed an unexpected market demand as companies downloaded it one after another. “What I had actually built was a routing and scheduling engine,” Stewart explains. “It evolved from pins on a map to a route optimization tool for salespeople who then wanted to use it for dispatch and more complex applications.”

However, getting in touch with Golden proved more difficult than getting a package from point A to point B. “This guy is practically nocturnal, so he’s hard to get a hold of,” he explains. After several false starts, Golden agreed to a phone call—at 1 a.m. As Stewart described the system he envisioned building, the management science professor became incredulous. “He said to me, ‘Do you have any idea how much money you’d need to do something like this?’” recalls Stewart, who responded that he had just raised $42 million in a second round of financing. Golden fell quiet before adding, “In that case, you’re going to need Dr. Chris Groer,” one of the top minds in the vehicle routing field.

When Stewart set about luring Groer and optimization expert Carnes from KMPG to tackle the routing and scheduling challenge at MapAnything, he appealed to Groer’s “just for fun” interest in another numbers problem: sports scheduling. Turns out Groer’s sister-in-law, who was head coach for a collegiate women’s soccer team, had complained to him that her team’s season was kicking off with five away games in a row, a taxing road trip for athletes of any stripe. Himself an athlete, Groer had started crunching the numbers at his kitchen table to see if he could find a better way of scheduling.

Although Groer had put his newfound hobby to work for his then-employer, KPMG, teaming up with Carnes to secure the scheduling contract with the NBA, Stewart enticed the duo to join MapAnything with a promise too good to refuse: After the core sales product was built and launched, the trio would work together to found a company focused only on sports scheduling, with the goal of taking the technology to the next level. Groer and Carnes joined the MapAnything team in 2017 and two years later, Salesforce bought MapAnything for $250 million. Although now poised to fulfill this long-promised dream, Stewart was faced with a significant, if temporary, barrier—a three-year noncompete agreement.

“It was pretty terrible,” says Stewart of the three years he spent laying low. “I sat on a lot of boards and invested in a lot of companies. But as coaches will tell you, they’d rather be out on the field.” Stewart used his forced hiatus to plan his next move, a venture that would reunite him with Groer and Carnes—and forever change how sports games are scheduled.

Fastbreak AI’s software currently relies on traditional machine learning, training itself according to inputted data, but requires clients to navigate complicated technical interfaces to set up the schedule’s rules. The company’s next software version will employ generative AI on the front end, allowing clients to provide input using everyday language, rather than creating rules through a series of “picks and clicks” in complex menus. In this way, the client will experience the Fastbreak AI interface much as they would ChatGPT and other large language models. “Any business user will be able to type in a prompt like, ‘I need St. John’s to play a game in Madison Square Garden on this date and broadcast on CBS,’ and we’ll validate that rule,” Stewart explains.

In less than three years, Fastbreak AI has become the dominant force in professional and collegiate sports scheduling. Yet Stewart’s vision extends beyond elite sports. “We plan to get deeper into youth sports, which is a $40 billion marketplace ripe for disruption,” he says, speaking as both CEO and a parent of five boys. “Right now, I’ve got six different apps on my phone for my kids who play travel sports. That’s not a great experience.”

The frustration parents experience extends beyond technology to the tournaments themselves. “They might play a game at 8 in the morning and then their next game is at 4 p.m., meaning you’re sitting in a sweaty gym for hours.” Fastbreak AI’s scheduling system would intelligently cluster games with just one to two hours between competitions—transforming the experience for young athletes and their families while opening new revenue opportunities in an underserved market.

Yet for Stewart, some measures of success transcend business metrics, as evidenced by a memorable interaction between his son and a Fastbreak AI investor. “My son Connor, who plays wide receiver, was walking by my office while I was on a video call with Larry Fitzgerald,” Stewart recalled, referencing the Arizona Cardinals’ former wide receiver. “I introduced them, and Larry asked if he was out there putting in the reps and running routes during the offseason. Connor made the mistake of saying no.” What followed was an impromptu pep talk from a College Football Hall of Fame player to an awestruck eighth grader.

The moment encapsulates Stewart’s unique position at the intersection of sports, technology, and opportunity. While he continues to revolutionize how leagues manage their most complex logistical challenges, the impact of his work touches something more fundamental—the experiences of athletes and fans at every level. From NBA superstars to youth league players like his own sons, Stewart’s algorithms are silently orchestrating the games that bring communities together, one optimized schedule at a time.