As the fourth of five brothers who’d spent a good portion of their teens fixing up old junkers, Keith knew how to retune the distributor with his trusty timing light.

But short trips remained a better bet than long ones—least of all a cross-country haul featuring a detour through Alabama.

Apart from his girlfriend Rene, the most important thing the Scirocco carried was a Visual 550 graphics terminal. Outfitted with a 14-inch monochrome screen whose phosphor pixels glowed green against a ghostly black background, and a graphics card that emulated the Tektronix 4014 storage tube, the Visual 550 existed close to the cutting edge of the burgeoning digital age. It was a $2,7001 machine (equivalent to $8,400 in 2024) that represented an intermediate step between massive punch- card mainframes and the PC revolution that IBM and Apple had only just begun to unleash. During an 18-month stint working for DuPont—his first job after earning a master’s in electrical engineering from the University of Florida—Keith had figured out how to use such a terminal to display CAD drawings that were otherwise only viewable on systems that sold for upwards of $70,000 per seat. The planned stop in Huntsville, Alabama, was to show his software to engineers at Intergraph, one of the “Big 5” CAD vendors of that era, which made the expensive hardware DuPont used for computer-aided drafting.

The Visual 550 was roughly the size of a standard cathode-ray-tube television set. After satisfying himself that it was packed securely among clothes suited to sunny Pasadena, Keith slipped his toolbox into the remaining space. “I knew this car wasn’t going to make it all the way across without lots of work,” he would recall. But at least it looked good from the outside. He and Barry had painted it themselves, in a glossy forest green. Now Barry had summoned Keith westward—to their mother’s dismay—to join a three-person software startup whose prospects were frankly no more certain than the Scirocco’s.

Barry and Keith were born two years apart, in the midst of a five-year stretch that saw Tom and Bonnie Bentley bring five sons into the world. Greg came first in 1955. Barry, Scott, and Keith followed at one-year intervals. Ray, born in 1960, would later quip that the Bentley boys “were like Irish quintuplets.” The quick succession of their births was only one aspect—perhaps not even the most important—of their closeness. “We always shared bedrooms, we always shared clothes,” Ray said. “Keith and I, people thought we were twins.”

Cindy came along as the sole sister in 1961, but the boys were especially tight. Growing up in the Hills of Skyline neighbourhood of Wilmington, they shared newspaper delivery routes and they worked for Wassam’s general store. They learned how to play band instruments at the behest of their mother, a former English teacher who also required them to read for at least an hour a day over the summer and produce book reports. They inherited the same emphasis on hard work and high standards from their father, who’d served four years in the Air Force before becoming a mechanical engineer at DuPont, and instilled in them an affinity for tinkering— allied with an inclination toward thrift.

“He would always have some project, and he would have one of his sons to help. So we were always involved in fixing stuff.” Keith recalled.

Assisting Tom Bentley in the garage could be a delicate business because he exuded a mix of traits common to many self-assured engineers. Curiosity and resourcefulness rubbed up against a sometimes daunting level of competence and exactitude.

“He was a very friendly person,” in Barry’s warm but wry recollection. “He was a natural leader, he volunteered for all kinds of things, and he was well-liked. But one thing you couldn’t say for him was that he was pleasant to work with— because he was not. He thought that if he didn’t do it himself, it had about a zero per cent chance of being done correctly. So if you were, like, helping him clean the basement, and he’d given you the task of sweeping the floor, he would criticize the vigor of your brush strokes. And he’d give you a few hints about how to hold the handle of the broom a little bit more effectively.”

From the vantage of adulthood, exercising a well-honed knack for familial self-deprecation, Barry conceded that when it came to him and his brothers, the apples didn’t fall far from the tree. “You might recognize some of those characteristics in some of us,” he suggested. All the same, the Bentley boys cottoned onto their father’s mechanical aptitude and charged it with a thrilling level of freedom and independence.

We just had one rule. You had to be home before dinner.

“We were typical boys,” said Keith. “We built go-karts and we had go-kart races. We had mini-bikes. I remember once that Barry and I had a go-kart and the chain broke, so we needed a master link. So we walked from our house in Delaware to the shopping center, which was probably five miles away, crossing highways. I was probably eleven!

“We just had one rule,” he remembered: “You had to be home before dinner.” Their pursuits matured as they entered their teens, but that sense of initiative only intensified—fostered continually by parents who conveyed faith and expectation in their children’s capabilities. When Greg obtained his driver’s license at 15, Tom Bentley decreed that his oldest son could exercise it upon fulfilling a series of conditions: he had to buy a used truck for less than $150, it couldn’t be running at the time of purchase, and he would be granted one year to make it road-worthy. Greg found a 1956 Ford pickup in an old field and talked its owner down to $25. He and Barry got it running within about six months, after which they used it to launch a trash-hauling business.“Once Greg started it with the truck, there was a stream of old cars that we took apart, rebuilt, painted over,” Keith said. By the time Ray came of age, his older brothers had sold the original truck, “but my father told me the same thing: get a truck from a field and rebuild it.” Ray found a 1952 Willys Jeep for $135 and revived it with Keith’s help—“though it wasn’t to restoration quality,” acknowledged the youngest Bentley brother, who dreamed of Mustangs and muscle cars.

Tom Bentley watched through clenched teeth as automotive paint slowly speckled the walls of his garage, but could only swell with pride at the determination and thriftiness that marked his sons’ growing self-sufficiency. He’d worked his own way through the University of Connecticut’s mechanical engineering program by supplementing GI Bill benefits with part-time jobs. Greg, Barry, and Scott had been born during his freshman, sophomore, and junior years. They’d lived on top of each other in a shoebox apartment. As Bonnie later recalled, “We thought it was wonderful when we were accepted in a federally funded low-income housing project—occupied by people on welfare and college students—in an area close to the college in his last year of school.” The house they drew may have been beaten up, but it gave Tom some quiet space in which to do his homework, and meant Bonnie “didn’t have to go down to a basement anymore with a laundry basket on one hip and a baby on the other.”

As parents, Tom and Bonnie strove to cultivate the same habits of hard work, love, dedication, and self-discipline through which they had improved their own station in life.

There appeared a terminal which was hooked up to something at the University of Delaware, which was a computer.

“Our family story probably would have been more interesting if we’d had some tremendous hardships to overcome while we were growing up,” Barry reflected. “But the only hardships I remember were our parents both pushing us pretty hard to get good grades in school, and my mother making us eat liver and beets.”

In the meantime, the boys stumbled into a domain considerably more exotic than combustion engines. Greg recalled his first glimpse of it in the library of John Dickinson High School, which the Bentley children attended. “There appeared a terminal which was hooked up to something at the University of Delaware, which was a computer,” he said. It had been installed by an initiative called Project DELTA, which aimed to bring minicomputers to public school students across the state, and Greg took to it at once.

His younger brothers quickly followed suit, taking advantage of a program whose timing could not have been more fortuitous. Inaugurated in 1971, Project DELTA (which stood for Delaware’s Total Approach to Computer Knowledge) amounted to a radical experiment in granting adolescents direct access to programmable computers well before the rise of mass-market PCs.2

“The University of Delaware had set up a PDP-11 computer,” Barry recalled. That 16-bit machine, which resided in the university’s electrical engineering department, was accessed by modems over conventional voice-grade phone lines. “They had set up 110-baud terminals at a bunch of local high schools, and Dickinson was one of them.

I saw the thing and I was fascinated. BASIC was the only language you could do, and you couldn’t store a program on the computer— you had to store it on paper tape.”

The first program Barry wrote on it was a blackjack game. “I figured out how to get a random number, and I figured out how to draw a card from that random number. And then I had the computer deal, and you could take a hit or not take a hit.” He did it all on his own. “I wasn’t taking a class in it or anything like that. All I could do was bumble my way through.”

“We have binary minds,” said Scott, who was, if anything, still more smitten. Even at the rate of ten characters per second, which was as fast as the signals were transmitted, there was nothing like programming. He created one program that summarized the high-school football team’s play-calling statistics and another that played and solved word puzzles. Come Thanksgiving and Christmas, he would lay a couch cushion in Greg’s 1956 Ford pickup, nestle the teletype terminal on top, and transfer it to the Bentleys’ basement for the holidays.

“We owe an awful lot, if not everything, to Project DELTA,” Scott reflected. “We were born at the right time. We got into computers back when it was a mystery. It was a black box to almost everybody.” But Delaware’s public- education initiative gave the Bentley brothers a jump on cracking it open.

Keith was similarly transfixed. “The teletype terminal looked like an automatic typewriter, and you could only type one line at a time but it was in the library at my high school, I found it, and I thought it was the greatest thing ever. I remember having to sign up for time, and I got Thursday morning from 8:00 to 9:30. Sometimes I would be there after school too, and occasionally I would get it over lunch or one of my study halls. It was more fun programming that computer than anything I had done before.” The first thing he created was a rudimentary stab at graphics, using BASIC to form images out of alphanumeric characters.

Yet it took a while for the magnetic pull of computing to fully overpower Keith’s first, rather quixotic, career fixation. Somehow, he and Ray had both concluded that their futures lay in dentistry. The youngest Bentley brother’s rationale was watertight. “They were the guys with the nice cars,” said Ray. “We had a Volkswagen bus because a three-row station wagon wasn’t big enough.”

The idea appealed to Keith for a slightly different, but revealing reason. Dentists, by and large, were self-employed. “I thought it would be great—you work for yourself, you set your own hours,” he explained. But by his own admission, he hadn’t thought much beyond the attraction of professional autonomy. “I didn’t quite realize that being a dentist meant putting your hands in other people’s mouths.”

After graduating from high school Keith enrolled in the University of Delaware’s pre-dentistry track, which shook him to his senses. “It was idiotic,” he laughed, reflecting on his freshman-year preconceptions. “But I took a couple of chemistry classes, which I hated—and then I took a physics class, and I said, ‘This is what I like: physics and math.’ And then I took an engineering class.” The lightbulb blazed. Keith changed his major to electrical engineering, and his path was set. He was well on the way to becoming his father’s son.

That was the pain point Keith figured out how to solve: writing software that permitted engineers to access their digital drawings on an ordinary graphics terminal that cost one-tenth the price of Intergraph’s workstations. He called it ViewDGN, after the DGN file format used by Intergraph’s Interactive Graphics Design System (IGDS), and pretty soon DuPont’s engineers were clamoring for it.

Though the implications may not have been fully apparent to anyone at the time, this was a small but critical catalyst of the process by which computer-aided drafting would be transformed into computer-aided design. Collapsing the boundary between engineers and CAD operators heralded a revolutionary change.

By his one-year anniversary at DuPont, Keith was proving his value to the company by any measure. His parents could hardly have been more content. Tom Bentley still worked there as a mechanical engineer, and Bonnie saw Delaware’s most prestigious firm for what it was: a Fortune 500 giant (#8 in 1983, with annual revenues exceeding $33 billion4) that all but guaranteed lifetime employment working at the forward edge of just about every field the company touched. “For my mother,” Keith said, “getting a job at DuPont was the pinnacle.”

But he couldn’t shake the feeling that he was in the wrong place, that this corporate behemoth was too big to make a meaningful impact on— or even to comprehend. “I was just one guy in a cubicle, working on something that was in support of something that was in support of something, in support of something,” he recalled. “I didn’t even know what we sold. I didn’t feel like I was contributing to its success—or if I was, I didn’t know how. So I thought that I’d really rather work for a smaller company.”

And as fortune had it, Barry had just started one. He’d banded together with two fellow graduate students at Caltech, where he had gone to pursue a Ph.D. in chemical engineering after earning a bachelor’s degree in that discipline from the University of Delaware in 1978. Studying fluid mechanics, he’d gravitated almost immediately to the intersection of chemical engineering and computing. “But I wasn’t the most enthusiastic graduate student,” Barry reflected. “I wasn’t super gung-ho research-oriented.” The summer before enrolling at Caltech he’d worked alongside Greg at a small consulting firm called the Yardley Group, which was involved in software development. It was a small shop whose entre- preneurial energy appealed to Barry, who learned a “very obscure but very cool” programming language called APL during his short stint there. That experience primed him to jump at the chance to join two peers in an effort to build a business around integrating laboratory instruments with the Apple II computer.

They founded a company called Dynamic Solutions, and set about combining custom-built hardware with software Barry helped to write for the Apple II “that would control and process the data from chromatography instruments, color imagery instruments, and things like that,” and display the output via bitmap graphics. They called their product “the Appligrator.” An early print advertisement emphasized its ability to produce graphics printouts with no programming experience required. “We took the APPLE and turned it into an intelligent laboratory integrator,” the ad declared. “We call it Appligration. You’ll call it Incredible.”

Before long two things became clear to the company’s founders: “The Apple II wasn’t good enough,” Barry said, and Apple as a company was moving away from the expandability it had originally offered via card slots that permitted users to customize their machines. That meant Dynamic Solutions needed to shift to IBM’s PC platform—for which they could use some programming help. So Barry reached out to Keith.

“The typical Intergraph system of the day cost about a million dollars,” Keith recalled. DuPont’s setup featured about 18 workstations, ranging from “cheap” $50,000 models to dual-color-monitor machines that approached $150,000. Keith gave a demo of ViewDGN to a group of Intergraph technical employees and managers.

“I thought it went pretty well. They were asking me all kinds of questions.” The fact that he’d written the software as a DuPont employee put him in a slightly odd position. DuPont technically owned ViewDGN, but as its sole creator, Keith was really the only person who knew how it worked. So he’d come to a handshake agreement with his boss before he left: in exchange for continuing to support and upgrade the software for DuPont, Keith would retain the ability to sell it. Intergraph, whose file format and operating architecture it was based on, was really the only logical buyer—and thus had all the leverage to dictate the terms.

“In my wildest dreams, I thought they might be willing to pay me $5,000 for it,” Keith remembered. But Intergraph didn’t know whether ViewDGN portended dreams or nightmares. His demo provoked curiosity— and something like an allergic reaction. The company was minting money on hardware sales. His software represented a silver opportunity that threatened the golden goose. “The ability to take a $5,000 terminal and use it for something that they probably would have sold you a $50,000 workstation to do—they weren’t sure whether they liked it or hated it,” Keith said. “In fact, they hated it. So they decided to do nothing.”

Keith returned to the seedy hotel where Rene was waiting for him, empty-handed. His strikeout at Intergraph became the first in a string of what Barry would later dub “fortuitous circumstances.”

“If they’d written me a check for $5,000,” Keith reflected, “Bentley Systems never would have existed.”

For Barry and Therese Bentley, who had married in 1982, Keith’s arrival augured a housing upgrade of sorts. They moved from an apartment to a two-bedroom house on the very edge of the upscale enclave of San Marino. “We could only afford it because we had Keith” pitching in for the monthly payment, Barry recollected. “And that got a little dicey,” he continued, “because Dynamic Solutions was not exactly a reliable payer.”

The fundamental problem was that the company’s technology was more promising than its market. The scientific laboratories that constituted their potential customers were habituated to analyzing data the old-fashioned way and tended to have dismal budgets for anything new. Dynamic Solutions sold their Appligrator for about $6,500—but that included a souped-up Apple II, and sales depended on independent dealers who took a big cut.

“The laboratory guys were scrimping to even try to get $6,500,” Barry recalled. “It was like pulling teeth.” Furthermore, lab managers required “a lot of handholding” to troubleshoot analog-digital interfaces and other challenges. “You had to match up the impedance correctly and you had to doink around with it a lot in order to get it to work,” he said. “It was a very frustrating business model.” As a result, the company was perpetually short of funds. “And one of the first things we did with funds we didn’t have,” Barry quipped, “was to not pay Keith.”

As he worked for intermittent wages, Keith continued to ply a parallel path. On October 11, 1983, he obtained a formal letter from DuPont permitting him to sell ViewDGN. Armed with a solid claim to the intellectual property, he pivoted to market research. Having attended an Intergraph users’ conference the year before, he’d held on to a master list of attendees. He sent some of them a letter explaining the software’s functionality along with a three-page questionnaire that aimed to gauge their interest. Question number nine out of 14 cut straight to the point: “If you were making a purchasing decision for the system as described: What is the minimum price you would expect to pay? What is the maximum price your organization would justify?”

After months of bouncing off the tightened purse strings of laboratory budgets, the response resounded like the blast of a trumpet fanfare. “People say that when you do blind mailings, you expect to get back five percent or less,” Keith reflected. He got 40. Better yet was their willingness to pay. “I just added all the responses up, averaged them, and it came out to something like $9,000.” And that was purely for the software. You didn’t need an MBA to recognize that the CAD market was a fatter target.

Barry—who had suggested renaming the software PseudoStation but had yet to actually see it run—“couldn’t believe it,” Keith said. So they decided to call a local respondent to offer a demo. They chose C.F. Braun, an engineering and construction company based in Alhambra, California, about five miles away from where the brothers now lived together.

For Barry, it was an unforgettable day. After being welcomed at C.F. Braun, the Bentley brothers were escorted through a darkened room whose edges glowed with the screens of eight or ten Intergraph workstations. In a room beyond that, Keith set up the Visual 550 to demonstrate PseudoStation to a manager named Dave Snyder. Barry fiddled around with cables and connections in an attempt to project an air of proficiency—something he rarely had to feign. “This was the first time I’d ever seen CAD,” he said. “I was trying to look busy.”

Keith had two goals for the meeting. The first was to determine whether anyone would truly consider buying this thing. The second was to convince Barry that their future might lie in CAD software. By the time they walked back out into the California sunshine, it looked like he’d hit the exacta.

“It was a revelation for me,” said Barry. “The demo went absolutely fabulously. Everything worked. You couldn’t get any better ... and I was thinking, ‘You know what, this could be a lot of fun!’” In fact, the meeting went so well that C.F. Braun invited the Bentleys back to present to a larger group. So about ten days later the brothers returned, riding high. The uncertainty that colored the first meeting had metamorphosed into beaming confidence. What followed was a scene Barry would relish recounting in the years ahead.

“It was a big, important demo. They had us in a nice conference room. Dave Snyder’s boss was there, and his boss’s boss and everybody looked important and had suits on. Keith and I—we also had suits on—had dragged in this terminal, hooked it up ... and proved the rule that the more important the demo, the more likely it is to be a complete, utter, unmitigated disaster.”

The culprit would only become evident too late. “The terminals were connected over an RS-232 line,” Keith recalled, “and the second one they gave us to connect to was misconfigured.” As it filtered out critical data that passed between the terminal and the VAX server, problems cascaded.

“We brought up the initial opening screen, showed the design file, and everything looked fine,” Barry remembered. “But we kept trying to zoom in, and it didn’t erase—all the vectors that came out when we zoomed in were painted over the screen again, so it was just a big mess. The only way we could get it to erase was to turn it off, wait about 5 seconds, turn it on again, and then zoom to the next level. It was terrible.”

Keith stole a glance at the Braun contingent’s senior man. He was staring at the debacle and rolling his eyes. On his way out, he uttered a phrase destined to become an unofficial motto in the years ahead, “an expression,”

I’d like to see it work a little better.

Barry would later reflect, “that I have heard thousands of times at Bentley Systems.” In a gem of sotto voce understatement, the man simply said, “I’d like to see it work a little better.” But the kicker was that he was willing to give the Bentley brothers a chance. “They actually bought it anyway,” Barry said. “And we were in the ViewDGN business.” Keith filed registration paperwork for Bentley Systems, Incorporated, on September 5, 1984,⁷ naming Greg as a member of the board.

He’d originally wanted to name the company Bentley Engineering, but that name was already taken. Yet in some respects, the name was

an afterthought; Keith expected that nobody would ever see or hear it, because the entity was initially intended just to collect royalties from Dynamic Solutions, which would sell the software pursuant to a one-year marketing agreement the two firms had signed in June 1984. By this time, Dynamic Solutions had a full-time salesperson, so the arrangement made sense.

The first order of business was to improve the software. It did need to “work a little better”— and the brothers also wanted to create versions that worked on other vendors’ terminals, beyond the Visual 550. Here they benefited from another fortuitous coincidence: Caltech’s chemical engineering department, where

Barry was still enrolled and had just purchased a DEC VAX minicomputer—which they could use to refine PseudoStation. “It was underutilized,” Barry recalled. “So they allowed students to buy time on it with their student account.”

Departmental overseers presumably intended the VAX to be used for academic purposes, not entrepreneurial ones, but it often sat idle in the wee hours when nobody was around. And that suited the brothers just fine.

“Keith was working full days on the Dynamic Solutions laboratory analysis stuff,” Barry explained. “And being a graduate student, I didn’t have any classes by that time but I was supposed to do my research. So we’d almost always work at night.”

“It was perfect,” Keith recalled. “We would sneak in at night and program on the Caltech VAX. And then we would sneak into the computer room, where the tape drive was, and we’d make tapes— because that’s how you distributed software back then.”

Starting around 9 p.m. and working deep into the witching hours, they’d emerge with a tangible demonstration of just how compelling a pure software business could be. “We used to load the cheapest blank tape we could find into the VAX drive,” Keith said, transferring the PseudoStation code onto it, “and say, ‘Hey, we just turned an $8 tape into a $9,000 tape!’”

In the first year, they sold approximately 80 copies of PseudoStation at an average price of $7,500. Life was good in San Marino, where the brothers grooved to Pink Floyd after splurging on a pair of stereo speakers made by a local audiophile outfit called Gross National Product. That fall, Barry and Therese welcomed their first son, Thomas, into the world. The revenue generated by Bentley Systems all but saved Dynamic Solutions, whose laboratory analysis product was having a harder time gaining market traction. But by the early months of 1985, the approaching expiration of the marketing agreement revealed a fissure within the latter company.

“Keith and I were on the side of, ‘Let’s invest in this market,’” Barry said. CAD operators had healthy computer-related budgets, and the increasing viability of IBM’s PC as a feasible alternative to proprietary graphics workstations held the possibility of an even more lucrative opportunity. But Barry’s partners didn’t see it the same way, and the differences were irreconcilable.

For Keith, the writing was on the wall. “Pseudo-Station was generating most of the revenue, and it was silly for me to be spending my time working on the software that was not contributing.” That meant that his time with Dynamic Solutions—both as an employer and as a marketing partner—had come to an end. Barry concurred. Indeed, the business case for Bentley Systems had become clear enough to him that Barry had made a momentous decision of his own—swapping half of his equity in Dynamic Solutions for a 50 percent stake in the company that bore his family name.

“In retrospect, it was a pretty good deal for Barry,” Keith would later laugh. “But on the other hand, Bentley Systems wouldn’t have existed if Barry hadn’t been involved. So, it worked out pretty well.”

That wasn’t the only deal they struck. Indepen- dence from Dynamic Solutions also meant geographic freedom—and Keith was eager to rejoin Rene back east. So, the boy who’d grown up resuscitating rattletrap jalopies bought a brand-new Toyota pickup truck—and swapped Barry the Scirocco for those sweet GNP stereo speakers. As for who got the better end of that bargain, the Scirocco gave up the ghost inside of six months—but Barry would hasten to add that “there was a washing machine involved, as well,” and that the new parents landed that suddenly indispensable appliance.

The brothers completed their California chapter by scrawling a business plan in longhand on notebook paper late one night in the spring of 1985. “We went out six quarters, made up some numbers for revenue, and made up some numbers for expenses,” recalled Barry, who planned to rejoin Keith after completing his Ph.D. “This was the flimsiest possible reason that you could possibly have to start a company, but that’s what we did.”

Then, Keith loaded up his Toyota and drove back across the United States. Awaiting him was a 28-year-old cum laude Wharton graduate who had secured office space in West Philadelphia upon becoming Bentley Systems’ first hire. His name was Scott Bentley. A new chapter was set to begin.