MechSE alum Adam Tilton (BSME 2010) recently announced the successful seed funding of his latest startup, Driver. Featured in TechCrunch, Driver’s generative AI-powered platform can both translate complex technical documentation into concise, actionable explanations and generate comprehensive documentation quickly for new products.

This is a boon for the semiconductor industry in particular, where technical documentation for a single chip can be thousands of pages long—resulting in long, tedious hours of study for engineers to understand how to implement the chip in a system or product.

Tilton had been immersed in the frustrating process of digesting and implementing information from lengthy technical documentation while working as the Senior Director of Connected Product Software at Nike, where he focused on building machine learning loops that would send sensor data from wearable products through cloud-based algorithms and deploy results back to the products.

“For large companies that have thousands of internal applications and long software history, you almost always have to figure out how to work with something new at every step along the way,” Tilton said.

Selecting particular microcontrollers, motion sensors, and other hardware for the system required extensive reading and evaluation. “We had to expand our scope to think about the firmware architecture and, in particular, how we were running sensors [in Nike’s devices],” Tilton recalled. “That led to needing to build complex infrastructure on the back end.”

The Bartonville, Illinois, native eventually left Chicago, and Nike, for Austin, Texas. He later reconnected with former collaborator Daniel Hensley, who was already in the area.  

Inspiration struck while Tilton and Hensley were collaborating on an embedded technology project for Levels that had similarly dense operations documentation. “We used large language models to help us extract relevant information so we could integrate the technology into the company’s infrastructure,” Tilton said. In this case, the language models weren’t writing software for a new system, but rather assisting in quickly understanding complexities of the existing system to create a product development action plan.

“That’s when we had the idea to start Driver,” Tilton said. (Hensley is a co-founder and the CTO.) Tilton and Hensley immediately began working on their new startup alongside third co-founder and CFO Jimmy Hugill.

“We’ve had wonderful engagement from customers,” Tilton said of the startup’s success so far. “We’ve had the opportunity to sign up customers both from semiconductor companies who have to produce documentation and manufacturers that integrate those components and need to quickly understand how to build products.”

As Driver continues to grow, Tilton anticipates building out the engineering team and scaling the business to accommodate more customers’ needs. “Over the next couple years, we’ll focus on increasing our engineering capacity across all dimensions, from design and back-end infrastructure to front-end implementation,” he said.

Tilton’s successful string of startups began with Rithmio, a data analytics software company focused on gesture recognition for wearable devices. As a PhD student, Tilton and his advisor, Professor Prashant Mehta, won the 2014 Cozad New Venture Challenge. “[Mehta] and I attended a technical conference where someone gave a presentation on wearable technology,” Tilton recalled. “I had an idea for how our research could apply to that problem—how wearable technology could be solved with our technology.

As an undergraduate student, Tilton had taken ME 340 (Dynamics of Mechanical Systems) and 360 (Signal Processing).

“I really enjoyed those classes,” he said. “They weren’t everybody’s favorite, but I found those concepts to be really interesting. I think that was an inflection point for me as an undergraduate, figuring out that I wanted to go into systems and signals instead of thermodynamics or fluids.”

Mehta first met Tilton when Tilton was a student in ME 360, which he was teaching at the time. “While at Illinois, [Tilton] was already a great mentor to several undergraduate students in engineering, computer science, and business school,” Mehta said, noting that Tilton has returned to the classroom since his graduation to share thoughts with students. “He is a natural entrepreneur [and] you cannot help but be impressed with his passion and boundless energy.”

Their collaboration began when Tilton joined Mehta’s research lab as an undergraduate student. Mehta also connected him with then-PhD candidate Robert Gregg, now an associate professor and the Associate Director of Robotics at the University of Michigan. Tilton also met Gregg’s collaborator and peer, Salvatore Candido—now the co-founder and CTP of EvolutionaryScale.

“My first project was a bipedal robot with [Gregg and Candido], and it was a lot of fun,” Tilton said. He also petitioned the Department of Electrical and Computer Engineering to allow him to take a couple 500-level courses, including linear systems theory. Between his research and advanced courses, Tilton was on a natural path to graduate school.

“I think graduate school was a very different experience than undergrad,” Tilton said. “You’re able to immerse in so many topics and concepts that you don’t have the time for [with a full-time job]. I look back on it very fondly and almost with a little sadness that it’s over.”

However, graduate school did not go as Tilton had intended. Having completed enough computer science coursework, he proposed a thesis topic and petitioned the Computer Science department (now the Thomas M. Siebel Center for Computer Science) for his master’s degree. Unfortunately, without being enrolled full-time in CS, this was not an option. He continued his doctoral studies in mechanical engineering until he won Cozad and moved to Chicago to run Rithmio.

Tilton sold Rithmio to Bosch Sensortec after three years and hoped to return to Champaign-Urbana at that point to finish out his PhD. However, this option now presented unexpected roadblocks as well. Mehta advised that, because Tilton’s former PhD work, for which he had also been an Illinois Innovation Award finalist, had culminated in Rithmio, he would need to begin a new topic for his PhD.

“He recognized that I had done excellent work, but noted that several other people had now contributed through [Rithmio],” Tilton said. “I couldn’t come back and write a dissertation about technology that had essentially been built by the company.”

Furthermore, Mehta recognized that Tilton’s heart lay in entrepreneurship. “He told me, ‘You’re not going to be a professor, you’re going to be an entrepreneur,’” Tilton recalled. “‘You should not come back to do research at Illinois—go do the things in industry that you’re meant to do.’ And that was really good advice.”

Still, having worked toward his PhD, Tilton would have liked to leave Illinois with the degree. He now jokes, “I spent seven years at the University of Illinois to get a bachelor’s degree in mechanical engineering, a degree that I’m very, very proud of.”

Mehta proved right about Tilton’s career path—his entrepreneurial efforts continued to be lucrative. In lieu of returning to the U of I, Tilton took a sales role at Ocient, where co-founder Chris Gladwin quickly became a mentor in entrepreneurship. He later left to co-found the machine learning startup Aktive, Inc, which was acquired by Nike in less than two years.

“[Tilton’s] career demonstrates the importance of the computational mindset that we are trying to bring to our curriculum in MechSE,” said Mehta, who praised Tilton’s affinity for building strong teams. “He is a rare breed of engineer who is equally at home working on a software problem or a business problem.”

For students interested in an entrepreneurial career path, Tilton advises persistence and resiliency. “Entrepreneurship is a journey of constant iteration and learning,” he said. “Build the habit of driving projects forward and learning continuously.”