Michael Rosenbaum and the trouble with a perfect résumé

Michael Rosenbaum once got a government job because the person hiring him had attended the same graduate program. He has told the story himself, which is a useful precaution for anyone preparing to criticize the way other people hire. The familiar institution, the reassuring connection, the little click of recognition: he had benefited from the machinery he would spend years trying to change.

“We want people who look like us,” he said in 2017. In a hiring discussion, resemblance can mean a shared university as easily as a shared face. The applicant feels promising because something about the applicant feels familiar. An employer calls that judgment. A person left outside the room might choose a different word.

Rosenbaum founded Catalyte in Baltimore in 2000. Its proposition gave employers a reason to reconsider the people they had passed over: find aptitude, teach the work, and put people on real software projects. For someone with an immaculate educational record, he had developed an inconvenient interest in the shortcomings of educational records.

The credentials were excellent. The question was awkward.

His own route passed through Harvard, the London School of Economics, and Harvard Law School. He studied Russian studies and government as an undergraduate, economics in London, then law. His early career included work in the Clinton administration and a clerkship with Diana Gribbon Motz on the federal appeals court. The conventional system had plenty of boxes to tick beside his name.

Long before he became a founder, he was a Maryland schoolboy who liked mathematics. He competed as a mathlete through high school and later credited the subject with helping him through his degrees. There is something pleasingly consistent about a former mathlete asking a room full of employers whether they have checked their working.

The question he took into business concerned the relationship between talent and access. Employers could see a degree, an earlier job, a polished explanation of an earlier job. Could they see the person who had never received the first opportunity? A résumé records what has happened. Rosenbaum wanted to know what might happen if somebody received training and a chance.

That distinction became a company. Baltimore supplied a place to test it. The city mattered as more than an address: a theory about overlooked workers becomes considerably more demanding when it must employ people, deliver software, and persuade customers to return. An elegant argument can survive a seminar. A software business also has deadlines.

A library gives the theory a deadline

One early example was refreshingly unglamorous. King County Library System needed help with its software. In a 2013 account, the company, then operating as Catalyst IT Services, offered development from Baltimore and Portland at prices competitive with offshore work. Rosenbaum described evaluating 1,000 applicants to hire 150, using analytics to find programmers the market undervalued.

The library had compared seven other providers. Price helped, but so did the promise of frequent improvements. The transition was untidy: its existing system was poorly documented, and the proposed two-week development cycles initially ran longer. Eventually, fixes and features arrived each Tuesday. The library’s problem gave the hiring philosophy a practical examination.

For a founder talking about opportunity, this is an instructive detail. The customer was buying functioning software. Whatever the social purpose of the hiring process, the Tuesday release still had to arrive. The commercial bargain made room for people whom familiar recruitment habits might have missed, then required the team to do the work.

Even Tinkertoys had an interview

The method took years to develop. In 2018, Rosenbaum described experiments that included asking people to assemble Tinkertoys and observing how quickly they recognized a mistaken initial assumption. The objective was to connect observable behavior with later engineering performance. The path toward that objective was less orderly than the phrase “predictive analytics” suggests.

He compared those early years to throwing spaghetti at a wall. Statistical significance took three or four years; consistent accuracy took longer. It is an unusually domestic description of a technical undertaking. Anyone expecting the founding algorithm to arrive complete, wearing a small academic gown, would have been disappointed.

The admission also gives the project some human proportion. Rosenbaum was describing repeated attempts, uncertain signals, and a need to learn from the results. Selecting people differently would not count for much if the selection failed to predict useful work. The ambition required patient measurement as well as an attractive premise.

“Frankly, for years we just threw spaghetti at a wall.”

Michael Rosenbaum, 2018

The retirement that did not happen

In 2016, Rosenbaum recruited Jacob Hsu to run Catalyte. Hsu had led a large international technology business and was considering a rather different next chapter: moving to Sweden with his family and publishing an architecture magazine. A visit to Catalyte interrupted the plan. The employees and their stories made him reconsider.

One of those stories belonged to Tim Reed. After leaving Morgan State, where tuition had become unaffordable, Reed encountered a Catalyte advertisement on Craigslist and suspected a scam. He completed the screening, trained for six months without paying for the instruction, and joined the company. By the time his story was told in 2021, he was a quality-assurance manager.

Reed’s initial suspicion is worth pausing over. An offer of training followed by a job sounded dubious to the person it was supposed to reach. The difficulty of opening a route into a profession includes persuading people that the route exists. A recruitment model can be unfamiliar on both sides of the interview table.

For Rosenbaum, bringing in Hsu meant handing over daily leadership of the company he had built. That is a meaningful career event for any founder. Here it also brought another operator into the experiment, somebody whose experience of building workforces could help carry the approach beyond the founder’s own efforts.

When Brad Pitt joined the sales department

Explaining the business became easier after Moneyball reached cinemas in 2011. Rosenbaum recalled giving prospective customers copies of the book to explain the search for undervalued ability. The movie furnished a more widely understood reference. Baseball had supplied the analogy; Hollywood helped people remember it. A Sony Pictures technology executive even gave the company an original film poster.

“Brad Pitt has had some positive effects in the world,” Rosenbaum observed in 2022. It is a modestly comic footnote to a career concerned with mathematical prediction. The concept needed data, but the sales conversation benefited from a film star. Even a business skeptical of familiar signals can appreciate a familiar story.

Arena, which emerged as a separate company in 2014, extended his work into hiring and career decisions across other kinds of jobs. Today its stated purpose includes connecting people with work and improving how organizations find, hire, and develop staff. Its leadership page lists Rosenbaum as founder, CEO, and chairman. The original question about missed potential has continued to occupy him.

The candidate became the product

In May 2021, he entered Maryland’s Democratic contest for governor. His campaign brought the business argument into public life: people without conventional credentials could succeed with training and support, and government should help make those routes available. By September, he was talking about the appetite for policy conversations he encountered around the state.

By November 30, he had ended the campaign and announced a return to the private sector. It had lasted about six months. His statement kept the ambition intact while abandoning the immediate vehicle. The campaign had sought a much larger audience for ideas he had spent years developing inside companies. The audience, and the task of reaching it, proved different.

His reflection the following August was more revealing than a tidy account of lessons learned. “In politics, however, the candidate is the product,” he wrote. He described taking months to find a way to talk about himself that felt authentic. Building an organization around a problem had been familiar work; asking voters to choose him required a different kind of persuasion.

He also proposed paying people while they learned new skills, with support for expenses such as transportation and child care. These were proposals, never accomplishments of a Rosenbaum administration. They made explicit a practical obstacle beneath his career-long argument: adults have bills during the interval between discovering an ability and being paid to use it.

Back to the people at the beginning

Recognition followed in a different forum. The Maryland Chamber of Commerce inducted Rosenbaum into its Business Hall of Fame in 2023. In April 2025, he delivered the keynote at Johns Hopkins’ HopStart venture competition, where student teams presented businesses of their own. A career that began by questioning access to opportunity now included addressing people at the beginning of theirs.

There is an appealing complication in that scene. Rosenbaum has benefited from universities, worked with them, and spoken to their aspiring founders. His business premise asks employers to widen their view beyond those institutions. The useful question is how much more a person might be able to do than the record currently shows.

That question has survived a leadership handover, a second company, and a short election campaign. It can be asked of a software applicant or a founder with three degrees. Rosenbaum’s own story began with an interviewer recognizing a familiar graduate program. Much of the work since has concerned the applicant for whom nobody in the room recognizes anything yet.