Building Software for Hardware's Highest-Stakes Moments

Revel founder and CEO Scott Morton knows what it takes to sit through a rocket launch’s final minute with his own software on the line.

"When you're at T-minus five minutes, you've got time," he says. "If something's going wrong, you can stop the launch. When you're at T-minus ten seconds, there's nothing you can do. Human-scale response is just too slow. You're basically hands off at that point."

For most of his nine years at SpaceX (opens in new tab), Scott spent that final minute on the console. He wrote the software that loaded propellant onto Falcon 9 and later helped design the control system for Starship, where he had the authority, and the responsibility, to stop a countdown. Everyone in the room could hear him on the mic. Everyone at the company could pull up the data feed.

"Either it's going to work or it won’t. If it doesn’t, you're going to be thinking about it for the rest of your life," he says. And if it does? "There's nothing that can really put into words how it feels when everything goes well. You had a big role. You were essential to the success. It's unbelievable."

Scott built Revel for exactly these kinds of high-stakes situations. Founded in 2024, Revel (opens in new tab) makes software that engineers use to test and control complex hardware (opens in new tab): rocket engines, satellites, jet engines, nuclear reactors, and more. It replaces tools that, by Scott's own evaluation, hadn't changed in a decade, which limited how fast engineers could test, how much control logic they could write, and ultimately how much they could innovate. Now, Revel is showing engineers that there’s a better way. The platform is used to run more than 200 test systems at Impulse Space (opens in new tab), the engine stands at Astro Mechanica (opens in new tab) and Orbital Operations (opens in new tab), and the nuclear microreactor at Radiant (opens in new tab), to name a few. As of October 2026, Revel has raised more than $180 million and reached a $1.25 billion valuation, with Felicis as its first investor (opens in new tab).

From Mechanical Engineering to Software

Scott didn't come to software directly. He grew up in Madison, Wisconsin, where as a child he often took apart the family vacuum cleaner for fun. His grandfather, a Navy veteran turned electrical engineer, had started a company that made power supplies and transformers, and Scott has said he always found that amazing. He studied mechanical engineering at the University of Minnesota and worked in a robotics lab, where he soon hit a wall. "I could design and machine everything to make the robot, but I could not write the code. And the code is what's really going to animate your creation. It's what makes it come to life."

He convinced his professors to sponsor him into the computer science department for a master's, a rare move for mechanical engineers. "I wanted to be able to do the entire thing, hardware and software, myself," recalls Scott. Nikolaos Papanikolopoulos, who ran the robotics lab, later said (opens in new tab) that Scott was "the classic example of why you should not treat students in a traditional way. You give him the resources and let him run."

Scott arrived at SpaceX in 2015 as an intern on the launch fluid systems team, working as a mechanical engineer. Interns at SpaceX weren't given intern-level work. He was tasked with manning the controls that open and close the valves. That level of trust increased when he converted to full-time.

"I was given just a ridiculous amount of responsibility very quickly," he says. "You're taking two chemicals from the pad that are kept far apart, and putting them into one container inside the rocket, separated by a thin sheet of aluminum." A failure doesn't just cost the vehicle. "You're risking your launch site, and if the launch site is no longer there, you have to rebuild it before you can fly again."

His first big launch was at Vandenberg Air Force Base in California. He remembers walking into the control room. "There was only one person over thirty, and it was the launch director. You look around, and everyone else is a bunch of kids that just graduated. You're like, what are we doing here? Do we really have the right group to do this?" But the flight went flawlessly. "We crushed it."

That changed how he thought about engineering. He'd spent college in research labs, where the output was a paper and the first step was a literature review. But when designing the propellant-loading controller, Scott says all of that went out the window. "I just had to make the simplest possible design I could, something I could monitor live and use to understand if the launch was going well or not." Knowing he'd be the one on console narrating the system's health to the launch director, with the whole company listening, sharpened him as an engineer: "It drilled into me what actually mattered to deliver the system and make it successful."

SpaceX also taught him how to read failure. When a test flight ended in a failure, Scott noted the discrepancy between internal and external perceptions of the event: "The CNN headline would be 'SpaceX fails,' but internally we knew that it was way more successful than we thought it would ever be." Over time, Scott learned that there were two kinds of failures. Neither was optimal, but one that met mission goals was acceptable. One that marked a regression or breakdown of a previously functioning system that was already working was not.

As Scott gained more say in the roles he took on, he continued to seek out high-stakes environments, even if the risks were higher. "There are places in the company that are pretty low risk, and there are places that are flying closer to the sun. I definitely gravitated toward the flying-close-to-the-sun roles."

But over the years, he discovered the wall he'd hit in the robotics lab was institutional. Falcon 9 had a system that let mechanical engineers write some of their own control logic, "but it was still pretty limiting." After moving into software engineering on Starship to help build a better control system, he saw the limits of the internal tool. "We had always struggled to resource it. I knew it could be way better, but we weren’t able to get there." Then he started thinking about everyone outside of SpaceX doing similar kinds of work, who felt the same pain: "What about the rest of the world?"

A Step Change for a Stagnant Market

Scott’s team at SpaceX had built their own tooling because they surveyed everything on the market and came up empty. "It was clear that those tools were not going to work for us." About five years later, Scott examined his options again. "What had changed? Nothing." But he knew that creating better software was possible because his team had done it. "I started to see that I could make more of an impact by starting Revel than I could continuing at SpaceX."

He was out of SpaceX by 2024 summer's end and wrote the first prototype of Revel himself. Though he isn't committing code anymore, he misses it. "It's total flow state, writing code and solving a problem. But I'm pretty practical about where my time should go now." That same year, Felicis met Scott and became Revel’s first investor. "No one has lived this problem more thoroughly than Scott," wrote Aydin Senkut (opens in new tab) and Jake Storm, "nor is anyone more zealous about fixing this broken system." Felicis invested again in the Series B in February 2026.

Recently, most of Scott’s time goes toward recruiting. Early Revel hires came out of the SpaceX alumni network, followed by hires from Anduril and Palantir. "I was able to cherry-pick the best folks who I knew would do really well," Scott says. Revel’s success is "a hundred percent due to the people that have joined. That's why we're where we are today. There's no question."

The New Hardware Test and Control Stack

Every part of Revel is built so engineers can test and control complex systems themselves. The hardware-agnostic platform connects instruments from multiple vendors into one ecosystem, enabling engineers to easily write control logic and run benchtop tests from anywhere. They’ve also expanded to a command and control offering that supports always-on, mission-critical operations at launch sites, reactors, and industrial facilities.

One of the most interesting things about the Revel platform is how code gets written. Before getting a master’s in software engineering, Scott had always been unable to write the control software for his systems independently. He didn’t want the next generation of hardware engineers to be held back by the same barrier, so he created RevelCode: a Python-like, high-performance programming language with built-in safety guarantees that any hardware engineer can understand and write.

This is meant to replace the graphical programming tools that dominate most hardware testing, which date back to the 1980s (opens in new tab). Scott has a name for them. "I call it programming with crayons. It undervalues what a mechanical engineer can do."

Scott designed RevelCode around how those engineers actually work. "A mechanical engineer can set up the system, configure how it's going to work, make the control interfaces and the views they want, and then go write the code. We've designed this language so that a mechanical engineer with very basic programming can do it. We even have technicians writing code. It's that easy."

Scott's pitch for RevelCode is blunt. "It's highly performant, highly accessible, and it's safe. We're not giving up performance. We’re enabling anyone to write it. And if it compiles, it cannot crash. You can still open the valve at the wrong time, for sure. But all the software mistakes are covered."

Asked whether it becomes a lock-in advantage over time, the way CUDA did for Nvidia, he doesn't hedge. "It already is. We may eventually open-source it. I think we'll have a whole community built around it."

Customers are surprised by how fast deployment moves. Orbital Operations, a rocket engine startup, had Revel integrated within 24 hours and ran 40 hotfires the next day without a software engineer on staff. Their test engineering intern built the stand's control interface in a few hours. And Impulse Space's engine stand setups went from 14 days to 8 hours. "It's SaaS-level speed of getting people on board," Scott says.

Everyone at Revel understands that a software crash can put a rocket, a reactor or the facility around it at risk. "They've all operated in the same environments, where software has to work."

Julia Vyborny, Impulse Space's Director of Propulsion, puts the difference this way: "Before Revel, I worried a lot about a test executing reliably and protecting our hardware. Today I can worry less about the software executing, and just worry about the hardware."

A Team of Builders

Scott believes building as ambitiously as Revel does should be fun. "If you're working this hard, you should enjoy it along the way. It's tough at times, it's high stakes, high stress. But you should be having fun. Otherwise, I think you're doing it wrong," says Scott.

His goal is to make sure his team preserves enough energy for the real crunch without discouraging people who genuinely want to spend their Saturdays noodling problems. "There will be times when I have to say, ‘Guys, we got ourselves into a tough situation, we have to get this done.’ But we always want to make sure people have gas in the tank to do that."

As the company grows, he’s continuing to seek out builders who are fundamentally motivated by Revel’s mission. And with a generational opportunity to accelerate modern technology advancement in hardware and physical systems, the team plays to win: "There's never been a software system that can span the smallest test systems up to an industrial-scale platform running an oil and gas refinery. We are winning against incumbents across all domains."

Revel is hiring across product, engineering, design, and more. See open roles here (opens in new tab).