John Reese and the unfinished business of the PDF

In 1968, John Reese was helping computers listen to the moon programme. An assembly-language programmer working on Apollo telemetry, he noticed that hours of sensor tape were being processed even when there was no signal to examine. The machine was faithfully doing work that did not need doing. Reese and his colleagues built sampling logic. They stopped spending computer time on silence.

He remembers receiving a cost-saving award worth more than his annual salary. There are worse ways for a young programmer to discover a professional interest. The episode also supplies a useful opening to a career that can otherwise seem to wander: banking machines, financial systems, résumés, school transcripts. Across those settings, Reese kept finding an expensive interval between having information and being able to use it.

His latest company, MyDocs, addresses that interval in college admissions. He had already helped move transcripts from envelopes to electronic delivery at Docufide, the business that became Parchment. Now he is concerned with the information inside the delivered document. A PDF can cross the country almost instantly and still leave somebody a long afternoon of interpretation.

When customers chose the machine

Before the transcripts came the cash machines. At Citibank in the 1970s, Reese worked on self-service delivery systems: banking at home, at petrol stations, and on city streets. He recalls a team that custom-built and deployed more than 500 ATMs around New York because the equipment it needed was not available off the shelf.

The detail he likes to remember is the queue. Customers sometimes waited for an ATM while a teller was free. To a banker counting transactions, that might look peculiar. To someone designing a service, it was a verdict. People had found something in the machine’s convenience that made the wait worthwhile.

That observation became part of his approach to products. The public description of his career includes home banking and pay-at-the-pump systems, followed by executive responsibilities at TRW. The settings varied; each involved deciding how technology could make a service easier to use. Reese would eventually carry the question into education, where an apparently ordinary document could require several institutions to cooperate.

“I love to eliminate elements that get in the way of doing our best work.”

John Reese

He describes himself as a product guy. That seems a sensible label for someone whose attention returns so often to the transaction itself. What must the customer do? Which steps help? Which merely reflect how an organisation has always arranged its work? A teller window and an admissions office have different purposes, but both can accumulate routines the person outside the counter has to navigate.

A transcript needs somewhere to go

The early Docufide chronology has more than one beginning. Mark Cohen and Jeffrey Harris founded the original company in March 2002. Docufide, Inc. was formed in June 2003, with Reese appointed CEO. His current biography identifies him as a founder of the business that became Parchment. Those details place his contribution in a founding team and an evolving company, rather than a solitary invention.

Its task was practical: get a student’s academic record from a school to a college securely. The early service still accommodated paper. Secure Transcript launched in November 2003, and the first live electronic school-to-college delivery followed in July 2004. A service built to change an administrative habit had to meet institutions where they were.

Indiana became the first statewide contract in August 2005. The initiative was intended to cover more than 65,000 seniors, over 500 public and private high schools, and more than 75 colleges and universities. Scale arrived through the work of connecting organisations. A transcript, after all, has limited ambition as a digital object unless somebody at the other end is ready to receive it.

By November 2009, Docufide reported serving more than 4,000 schools and holding ten statewide contracts. Kansas had joined a roster that included Indiana, Michigan, Wisconsin, Minnesota, Nebraska, South Carolina, Utah, Alaska, and Arizona. The company said it had doubled its number of sending schools that year. It was also delivering electronically to the Common Application organisation’s 391 member institutions.

These are administrative numbers, which can sound unpromising until one considers the people behind them. Each school represented staff, existing systems, and decisions about accepting a different process. Each receiver made the network more useful to a sender. Reese was running a business whose usefulness depended on convincing people on both sides of the exchange.

In 2011, Matthew Pittinsky, a Blackboard co-founder, became CEO, and Docufide became Parchment. Reese’s subsequent roles included executive chairman and director. The change brought another entrepreneurial career into the company’s history. Pittinsky had encountered credential problems through his earlier work and saw room to build beyond the existing transcript service.

The business’s first-year total had been 128 credentials. It took seven years to reach one million. In February 2024, when Instructure completed its acquisition, Parchment had exchanged more than 165 million credentials and served over 13,000 customers. The later scale belongs to a company built over two decades by many people. Reese’s place in the story is the early work of getting the exchange established.

The information still inside the picture

A successful delivery system can expose the next problem. Once a transcript reaches admissions, a person may still have to interpret its layout, understand its grading scale, and turn its contents into information the institution can use. Electronic arrival does not settle those questions. The envelope has disappeared; the clerical work can be remarkably persistent.

Reese returned to that problem with MyDocs. The company applies optical character recognition and AI to transcript images, with the aim of producing structured academic profiles. Its proposition is that records from different formats, languages, and educational systems can become usable inputs for admissions work. The important step in that proposition happens after the file has arrived.

The company’s product names describe two points in the process. MAPit turns transcripts into MyDocs Academic Profiles. FirstLook allows students to upload records earlier in their contact with an institution, so recruiting and admissions teams can work with academic information sooner. These are proposed uses of the information, rather than a claim that the software should choose the student’s future.

In November 2024, Educational Credential Evaluators announced a collaboration with MyDocs around MAPit technology. The connection gave Reese’s new venture a partner already working on the interpretation of international qualifications. It also made the problem concrete: an academic record has to remain intelligible when it crosses an educational border, as well as a digital one.

Self-service has to earn its name

Reese has a specific objection to a certain kind of convenience. Asking applicants to type all their courses and grades into an online form can save an institution work while giving the student a new chore. His writing calls this task shifting. A service can look efficient from behind the counter while requiring considerable labour from the person using it.

He prefers a sequence he had explored earlier with résumés and employment applications: extract the information first, then ask the person to review what the system has done. The difference is visible in the blank form. One workflow hands the customer the entire job; the other arrives with work already completed and something specific to check.

In September 2025, he described FirstLook’s student confirmation feature. A student can review, edit, and confirm the academic profile created from their transcript. Changes are logged, and the original transcript stays unchanged. The design acknowledges that an extracted record may need correcting while keeping the underlying document intact.

That small piece of product design carries much of his philosophy. Automation takes on processing. A person gets a chance to inspect their own information. The institution retains its decision. The roles are different, and the workflow has to make room for each of them. It is the sort of detail easily lost in a conversation that begins and ends with the letters AI.

Another conversation, another company

Reese’s LinkedIn introduction calls him a “Non-recovered entrepreneur.” His return to academic records supplies the evidence. In February 2026, he wrote about wanting comparable, structured academic information while preserving each institution’s authority over evaluation. Colleges would decide which criteria mattered; the infrastructure would make the information easier to work with.

There is a sociable side to this long interest in removing unnecessary steps. Reflecting on the 2026 ASU+GSV Summit, Reese recalled pitching Docufide at the first gathering in 2010, when he remembered roughly 300 attendees. He praised the value of short, in-person conversations and said the later event had produced more than a dozen meaningful discussions. Someone interested in self-service can still appreciate a roomful of people.

The Apollo tape offers a fitting way to understand his persistence. The early improvement came from noticing how much work a computer was doing without a useful result. Decades later, the transcript has travelled electronically, but somebody still has to get its contents ready for a decision. Reese has found another interval to work on. This time, the silence is inside the PDF.