Art Papas started Bullhorn with a familiar kind of confidence: the confidence of a smart person who had found an elegant answer before meeting the question. In 1999, he and his co-founders imagined an online marketplace where creative freelancers could display their work and find contracts. Papas wrote the software. Prospective users looked at it. The applause was, at best, exquisitely polite.
The founders had not lived the problem. Nobody had cornered them and demanded this marketplace. They had assembled an idea because the parts sounded attractive together. Papas later described the market's puzzled response with comic economy: people kept asking how they had come up with it. The subtext was clear enough. The product was a handsome key designed without troubling to locate a lock.
A less useful founder might have polished the key. Papas eventually put it down.
A customer arrives with the plot
The decisive conversation was not about creative portfolios. It was with Leslie McIntyre, an early customer running a staffing business. Her teams were wrestling with cumbersome processes and the difficulty of coordinating work across branch offices. Papas proposed hosting the system on the internet, a wager that sounded riskier in the early 2000s than it does now. McIntyre trusted the proposal. Bullhorn acquired an actual job.
The distinction changed everything. Staffing firms did not need a prettier place for talent to advertise itself. They needed memory, coordination and momentum: who had spoken to which candidate, what a client required, where a placement stood, what action came next. Papas turned those accumulated frictions into a web-based system that let separated offices behave like one organization.
A narrow customer is not a small opportunity when the workflow is deep, frequent and expensive.
The first line of Bullhorn code was written in Papas's Cambridge apartment. At times, five people arrived there to work. He wrote much of the original applicant tracking system himself, and parts of that code reportedly remained in use until 2015. There is something pleasingly unfashionable about this artifact. Software companies are expected to replace their foundations with ceremonial regularity. Bullhorn kept a few apartment floorboards for sixteen years because they still held.
The market that looked narrow from outside proved roomy from within. A staffing company is part sales operation, part relationship network, part compliance machine and part financial plumbing. Once Bullhorn committed to that world, it could keep moving through the workflow rather than shopping for a new identity. Today its software covers the recruitment lifecycle from sourcing and customer management through onboarding, time collection, payroll and billing. More than 10,000 companies use it.
The engineer learns the CEO's job
Papas did not grow up announcing that he would become a software founder. His parents worked in medicine, and he entered Tufts University on a pre-med track. Then an optional calculus course caught him. He liked mathematics when other students regarded enjoyment as a suspicious response. The pleasure was in problem-solving, and it pulled him toward a mathematics degree, earned summa cum laude, and then software engineering at Thomson Reuters.
That background made him Bullhorn's original architect and first chief technology officer. Becoming CEO demanded a different kind of calculation. Code is obligingly explicit when it fails. Organizations can malfunction while producing attractive charts. Papas had to learn finance, hiring, culture, sales, capital and, later, acquisitions. The company accepted venture backing, moved through private-equity sponsors and bought other businesses. His title became more stable as the work beneath it kept changing.
His longevity is unusual because founder and professional CEO are often treated as separate species. Papas has been required to be both, sometimes in the same meeting. When discussing private equity, he has argued that a management team still needs to supply the vision. Investors can challenge assumptions, arrange financing and expose holes in a plan. They cannot usefully become the imagination department. Control, in his telling, is less a legal possession than the influence earned by having a credible destination.
When growth became a bad audience
Bullhorn's hardest pivot may not have been the first one. Early customers received intense attention because every relationship mattered to survival. Then the company grew, the technology was praised and sales accelerated. Papas has said Bullhorn began believing its own press. Its mission, once phrased around powering staffing desks, was jokingly shortened inside the company to “global domination.” The joke escaped its enclosure. Customer satisfaction fell and customers left.
Papas reset the mission around creating an incredible customer experience. The phrase could have dissolved into the beige wallpaper of corporate values, so Bullhorn attached it to visible behavior. At monthly town halls, employees were recognized for interventions that helped customers. Career growth and appreciation became part of the operating argument. Papas's compact view was that people wanted more than a paycheck; they wanted recognition and membership in something larger.
The useful point is not that a company should print “customer” in large type. It is that attention follows rewards. If the person who rescues a difficult account gets the microphone, colleagues learn what counts. If only the person who closes a new deal gets it, they learn that too. Culture is an editorial decision repeated until it looks natural.
The episode also reveals a recurring Papas trait: he is willing to make his own embarrassment instructional. The failed marketplace, the swaggering slogan and the dissatisfied customers all appear in his public telling. These are not charming mishaps pasted onto an inevitable ascent. Each one forced a change in what Bullhorn built or valued. Candor becomes useful when it is connected to a new operating choice.
The AI question is really a work question
By 2026, Bullhorn's frontier had moved from putting staffing records on the web to putting AI inside the work itself. At Engage Boston, Papas pushed against the idea that efficiency must end in fewer people. Bullhorn's product teams had seen work that once took weeks collapse into hours. Instead of treating the saved time as an invitation to shrink, the company pulled distant roadmap items forward and hired more product talent because additional ambition had become economical.
He pointed to a similar pattern among early customers using Bullhorn Amplify. Routine tasks could be automated, but the interesting result arrived afterward: recruiters spent more time building relationships, advising clients and pursuing work that had sat neglected. Bullhorn reported that these early users averaged 40 percent more placements per employee, with some reaching gains as high as 80 percent. The numbers belong to a defined customer group, not a promise to everyone. The principle is broader: cheaper effort can expand appetite.
That principle echoes the original pivot. The internet did not remove the staffing firm; it gave branches a shared memory. AI, in Papas's current thesis, should not merely imitate a recruiter at lower cost. It should turn a system that stores yesterday's activity into one that helps organize today's next move. Bullhorn's new digital workers focus on activities such as prospecting, verification, auditing and transcription. The nouns are prosaic. Good enterprise software often is. Nobody buys a payroll system for the poetry.
The founder who keeps changing jobs
Papas remains closely identified with Boston and Tufts. He serves on the university's computer science external advisory board. He has also volunteered with Career Collaborative, a Boston nonprofit that helps people build interviewing skills and move into more secure work. Even his recreational cameo fits the company-town quality of his story: at a 2017 Bullhorn Engage party, he appeared onstage with his band. A staffing-software conference contains multitudes.
His favorite interview prompts are almost disarmingly plain: what do you enjoy doing professionally, and what do you dislike? Candidates tend to reveal whether their appetites match the role. Someone interviewing for accounting once volunteered a hatred of spreadsheets. Papas appreciated the disclosure, presumably more than he would have appreciated discovering it after orientation.
There is a clean thread through these scenes. The mathematics student liked problems. The engineer built tools. The founder learned that a real customer's ungainly problem was more valuable than his own elegant premise. The CEO learned that growth could pull attention away from the people paying for it. The buyer learned that capital works best when management still knows where it is going. Now the AI advocate is asking what people can attempt when repetitive work recedes.
Bullhorn's endurance is not evidence that Papas had the right idea in 1999. He did not, and he says so. It is evidence that being wrong can be productive when a founder is more loyal to the problem than to the pitch. Twenty-six years later, Papas is still at the company because he has repeatedly allowed the job, the customer and the technology to revise him. The first plan did not survive. The habit of listening did.