This former hacker saw the light—and now wants to collect all of it

To control the light, Roelker cofounded a company in 2025 called Observable Space. He struggled to explain how he wound up here. His parents didn’t go to college. His dad died when Roelker was fairly young. His mom cooked school lunches. When Roelker left this working-class life in small-town Pennsylvania to attend a private Seventh-day Adventist university in Maryland, his family was stunned when he said he would study mathematics and philosophy. “My family was like, ‘What are you going to do with that? How are you going to be able to build a house with that?’” he recalled. But beyond his formal classes, what intrigued Roelker most were computers, particularly hacking. This was the late 1990s, and he had grown up with books like The Anarchist Cookbook and realized the power of computers and the growing influence of the Internet on society. Even before graduating, he took a research job at the nearby Johns Hopkins Applied Physics Laboratory. He could envision a four-decade career stretching out before him, providing the kind of financial stability he had not enjoyed growing up, a comfortable retirement with a public school’s 403(b) plan, and more. But this wasn’t enough. He sought adventure and excitement and, amid the dot-com boom, a chance at greater riches. So he left academia for the private industry, working as a software engineer at a small networking company before becoming a founding developer at Sourcefire in May 2002, a startup focused on network security. It would later be acquired by Cisco for $2.7 billion. Seeking to move from cyber defense to offense, he took on cyberwarfare jobs at companies later acquired by BAE Systems and Raytheon. These activities brought Roelker to the attention of the US government. “It was a pretty big deal because it was one of the first public acknowledgements that the military was engaging in offensive cyberwarfare,” Roelker said. After three years at DARPA and more than a decade in cyberspace and hacking, Roelker was ready for a reset. At the beginning of 2014, Roelker left DARPA to try something rather different, working on the most popular PC game in the world. At the time, League of Legends had an enormous global player base but faced problems with the “client” used to join games. He moved across the country to the Los Angeles area to work on this software. He liked the game (Roelker played mid, with Diana as a main), but found the workplace too laid-back, since its developer, Riot Games, was swimming in cash. So about a year after joining developer Riot, Roelker began looking around for something with a greater urgency. He’d been into space in his younger days and was an avid sci-fi reader, but he felt the field didn’t pay well. By 2015, SpaceX was starting to do some pretty interesting things with launch, however, and he lived near the company’s headquarters in Hawthorne. He made inquiries through connected friends and found out that the company needed someone to lead software engineering. The hiring process at SpaceX is brutal; for software positions, it includes a six-hour coding test. But Roelker got through and was excited to join in September. Just a couple of months earlier, the company had blown up a Falcon 9 rocket for the first time, and it was all hands on deck to get the rocket flying again. But SpaceX founder Elon Musk didn’t just want to return to flight; he also wanted to do so with a significantly upgraded version of the booster (using densified propellant) and to attempt a land-based landing of the first stage for the first time. On his first day on the job, Roelker met the Falcon 9 software team as their new manager. He was told the team was eight months behind schedule and represented the long pole for the vehicle’s return to flight, which Musk wanted to achieve before the end of the year. Half of the rocket’s software team was on the brink of quitting. “Good luck,” he was told. The new vice president of software engineering took up the challenge, seeking to pare back the department’s scope to focus solely on Falcon 9 flight software and the first stage landing. The rocket ultimately returned to flight successfully on December 21, and the first stage did indeed land. After, Roelker earned the confidence of Musk and a key lieutenant, Mark Juncosa, who began entrusting him with increasingly important projects, including flight software for Falcon Heavy, Crew Dragon, early versions of Starship, and, of course, the Starlink Internet constellation. By the late 2010s, work on the Internet constellation became all-consuming. “It was all in on Starlink because we had to get that revenue-generating engine going,” Roelker said. During this time, Roelker learned all about the space industry, the importance of vertical integration, and the many technical challenges involved in launching and managing hundreds, and eventually thousands, of satellites. One of SpaceX’s greatest challenges, he realized, was simply knowing the precise location of both its satellites and everything else in orbit. Every day, the Air Force would send a list of about a thousand potential collisions, but most were false positives. As the Starlink work piled up, Roelker left SpaceX in 2019. He was more interested in space exploration than building a massive telecommunications network in orbit. For a couple of years, he dabbled in crypto and emerging areas like non-fungible tokens. He was one of the first employees at OpenSea, leading engineering there. The company was printing money, and in January 2022, it raised $300 million at a $13.3 billion valuation. “I cashed out after that, and then crypto tanked,” Roelker said. After the cryptocurrency side quest, he was ready to get back into space. And he knew the problem he wanted to solve. The company’s instruments are useful to astronomers and are often found in remote observatories like Starfront. Additionally, in January, former Google chief executive Eric Schmidt announced plans to fund several large observatories, including an array of 1,200 telescopes, each with 11-inch mirrors, to mimic the effect of an 8-meter optical telescope. It will image the entire Northern Hemisphere sky. Observable Space won the contract to build all 1,200 telescopes for this project, called Argus Array. But for all that, astronomical telescopes remain a fairly niche industry. “Everything we know about the universe comes from telescopes,” Roelker said. “So while the total addressable market of that is not super big or attractive to VCs, I don’t know of a bigger question we can answer as humans outside of that. So we kind of want to have it both ways. We want to build this amazing technology that helps us understand the Universe, but we also know that same technology can be used for extremely big markets that are powering the space ecosystem.” The ways in which the US military tracks satellites are changing quickly, largely in response to a rapidly growing number of satellites. Over the last half-decade, the number of active satellites has grown from about 3,000 in low-Earth orbit to more than 15,000, and the number continues to grow with each launch of Starlink satellites, Amazon Leo satellites, and China’s Guowang and Qianfan constellations. This explosive growth has led the US military to increasingly rely on commercial data and to integrate it directly into routine military operations. The Space Force has also been seeking to acquire more ground-based sensors of its own to maintain coverage of crowded orbits. Last month, the US Space Force awarded a $94 million contract to Observable Space to expand its production of high-performance optical telescopes. “The Department is acting on the urgent need for mobile, off-grid robotic telescopes,” said Jeremy Verbout, assistant secretary for mission capabilities, in a statement. “These systems will provide the Joint Force with high-fidelity space domain awareness.” Observable Space is also developing new, lower-cost optics for use on spacecraft in orbit. Its Iguana space telescope, with a 200 mm aperture, allows operators to perform space domain awareness and astronomical observations and navigate for rendezvous and proximity operations. The first of these Iguana telescopes is set to fly on a spacecraft bus built by Apex later this year as part of its Project Shadow mission to demonstrate space-based interceptors. At a high level, the concept is fairly straightforward. A spacecraft encodes data onto a laser, which sends a narrow beam toward Earth. Large optical telescopes on the ground collect the incoming photons, and detectors convert the light back into electrical signals. Sophisticated error-correction software reconstructs the original message as many photons are lost. The greater the distance, the more daunting the challenge. A laser beam from geostationary orbit, about 22,000 miles (36,000 km) above Earth, starts out about the diameter of a coffee cup, and when it reaches Earth, it’s about 1km across. The farther away, the farther the beam spread, so ground-based telescopes can capture only a tiny fraction of the signal from distant spacecraft. The future of communication throughout the heavens will therefore probably be based on relay spacecraft, which are essentially like Internet routers here on Earth. “We’re going to be building the systems that get deployed in space and become the fiber optics infrastructure of communication across the Solar System,” Roelker said. But it is starting to happen. Observable Space played a key role in facilitating optical communications on Artemis II in April as it flew around the Moon. This type of high-bandwidth communications is expected to become standard for future Artemis missions and will enable lunar landings in high definition.