Retirement can be a surprisingly inefficient way to stop working. In March 2022, Paul Graziani left the chief executive’s job at Analytical Graphics, Inc., the company he had helped start 33 years earlier. His continuing assignment was another chief executive’s job. At COMSPOC, the assignment involved keeping track of things moving through space. Apparently, leaving one orbit was sufficient.
There is a pleasing consistency to this career. Graziani has spent decades arguing that people should have better tools for understanding complicated missions. First came software that engineers could buy and reuse. Later came a company concerned with the whereabouts and behavior of satellites and debris. The scenery expanded; the practical question endured. What does a person need to know before deciding what to do?
A biology degree, then a living room
His route into aerospace began with a degree in biology from La Salle College. Software development supplied the turn in the road. In 1980, he joined General Electric’s Space Division, where he held software and engineering roles. A decade that began inside a large aerospace organization ended with a much smaller workplace: his own living room.
In 1989, Graziani and two colleagues launched AGI there. The founders were Graziani, Scott Reynolds and Jim Poland. They came out of GE’s aerospace business with an idea that sounds almost modest when removed from its setting. Engineers working on different missions were repeatedly paying to solve similar problems. A reusable commercial product could spare them some of that repetition. The expensive habit was the opportunity.
Satellite Tool Kit, later called Systems Tool Kit, became the foundation. Even its name offers a clue to the thinking. A toolkit is something you use to build other things. The customer’s mission could remain particular; the underlying calculations did not have to begin again with every contract. Graziani’s business would depend on persuading technically demanding customers to accept that distinction.
The case against starting over
By 2004, Graziani could explain the argument in the language of a government budget. In an interview about military aerospace, he described AGI’s products as commercial software for planning and using spacecraft, aircraft, ships and ground vehicles. His examples included communications analysis, navigation and visualizing an operational environment. These were working tools for people with specific questions to answer.
He also gave striking estimates of the cost of government software development and the savings customers reported from buying existing code. Those figures belonged to that interview and its moment. The underlying proposition was simpler and more durable: spend scarce development money on the capabilities unique to your mission, while purchasing the foundation you share with everyone else.
That argument required more than a persuasive demonstration. Aerospace customers needed confidence that a product could support real work over time. Commercial software had to earn its place in an environment accustomed to commissioned systems. Graziani was selling a different allocation of effort, with engineers free to concentrate further up the problem. There is a certain courtesy in a business built around giving skilled people fewer things to redo.
Room for people, and for other companies
The organization around the software became part of the story. AGI received workplace recognition, including a place among The Wall Street Journal’s Top Small Workplaces in 2009. Its awards record also lists Graziani’s 2012 Aviation Week and Space Technology Laureate Award in the workforce category. These were acknowledgments of how the company operated as well as what it made.
The business reached beyond its Pennsylvania beginnings. In 2010, it received a U.S. Department of Commerce Export Achievement Award. At the time, exports accounted for 10 percent of revenue, and its internationally focused full-time positions had risen from six to eleven over two years. For a company selling specialized aerospace software, crossing borders involved building relationships and distribution, alongside writing code.
Graziani’s professional circle crossed institutional borders too. Robert Cardillo, the former director of the National Geospatial-Intelligence Agency, joined AGI’s board in 2019. Graziani welcomed the operational experience he would bring. The connection fits a recurring concern in the business: useful software needs a close understanding of the people who will rely on its answers.
Some ideas developed inside AGI eventually needed companies of their own. Cesium pursued 3D geospatial technology. OneSky worked on unmanned traffic management. COMSPOC concentrated on understanding activity in space. Growing a company, in this case, included allowing pieces of it to become independent. The family tree acquired several addresses.
A sale with three loose ends
In October 2020, Ansys announced an agreement to acquire AGI for $700 million. The attraction was mission-level simulation: connecting the behavior of components to the larger undertaking they served. A satellite’s equipment matters, but so does its relationship with a ground station as the spacecraft moves. The acquisition joined those scales of analysis within a larger simulation business.
The deal closed on December 1, 2020. Ansys’s subsequent annual report recorded a $720.1 million purchase price including net working capital adjustments. The announced figure and the later accounting figure describe different stages of the same transaction. Neither figure is a measure of Graziani’s personal wealth.
Cesium, OneSky and COMSPOC remained separate from the sale. A December 2020 account described Graziani’s ambition for a billion-dollar business and the possibilities still open through those independent companies. It also followed Cesium’s move into Philadelphia under Patrick Cozzi, whose team had begun the project within AGI. For Graziani, selling the established company left several unfinished interests very much alive.
The traffic problem above the traffic
COMSPOC’s roots reach back into AGI. The Commercial Space Operations Center was created there in 2014, and COMSPOC Corporation became independent in 2020. Its work concerns space situational awareness: maintaining an understanding of objects in orbit so operators can judge what those objects might do next. The difficulty lies partly in turning scattered observations into a useful picture.
Graziani brought that concern to a Senate hearing in July 2021. His prepared testimony criticized years of study, debate and insufficient action on space traffic management. He asked for clearer legislative direction and the use of commercial capabilities. Satellites support communications, navigation, weather information and commerce. Keeping their operating environment usable has consequences far below the altitude of the machinery.
The frustration sounds familiar. Early in his career, the obstacle was repeated software development. Here it was the distance between available capability and its adoption. Institutions have their own momentum. An object in orbit is famously difficult to slow down; a procurement process can exhibit an equally impressive attachment to its current course.
In April 2022, COMSPOC and ThothX announced a project to adapt a former Gemini and Apollo tracking dish in Western Australia for deep-space radar. The planned system would extend observations over the Asia-Pacific region. Old space infrastructure was being proposed for a new monitoring role: a useful reminder that progress sometimes involves finding another job for a very large piece of existing equipment.
A useful degree of doubt
Graziani’s explanations of the work leave room for uncertainty. In a 2024 interview, he described combining optical, radio-frequency and radar observations, while checking for inconsistent measurements. COMSPOC’s team nicknamed its Kalman-filter capability a lie detector. A sensor can be miscalibrated; a clock can drift. A disagreement in the data is something to investigate.
“This is an imperfect science,” he said. That admission matters when a prediction might inform a collision-avoidance decision. More observations need interpretation, and the reliability of a position estimate matters alongside the estimate itself. The operator needs a usable account of uncertainty, rather than a screen that merely looks certain.
Still looking for the useful answer
In a 2026 Space Symposium conversation with Laura Todd, Graziani returned to the burden on the operator. COMSPOC’s account of the interview described a growing tracked-object population and the challenge of recognizing the objects that deserve attention. He framed the problem around knowing what actually needs attention. It is a software requirement expressed in ordinary English.
“How can I see what I really need to worry about?”Paul Graziani · Space Symposium interview, 2026
There are other interests in the picture. Graziani is a woodworker, a sports-car enthusiast and an early adopter of technology. Those details make a welcome counterweight to the institutional acronyms. Beyond work, there are materials to shape and machines to enjoy. His degree in biology remains another reminder that a career need not follow the subject printed on the diploma.
The line connecting his businesses is the usefulness of an answer. AGI asked how much effort an engineer could save by starting with a shared toolkit. COMSPOC asks how an operator can make sense of a crowded orbital environment. Graziani’s version of retirement has kept him close to that question. Space may be enormous, but the task keeps coming back to the person at the screen.