Jennifer Smith kept finding the same contradiction at work. An employee would have a substantial training binder, the sort of object that reassures an organization that its procedures are under control. Then the employee would explain the shortcuts that made the job easier. The official instructions had weight. The useful instructions had an owner.
At McKinsey, Smith helped operations centers improve their performance. She would shadow a capable agent, learn the better way of doing a task, and help put those discoveries into PowerPoint for the client. In one telling, the agent had found 30 shortcuts. The company already employed the person with the answers. It was hiring consultants to help the answers travel.
There is a small comedy in that arrangement: the person closest to the work supplies the insight, someone else supplies the slides, and the institution buys back a tidier version of its own intelligence. Smith had a practical objection. Why should the employee need an intermediary to share what they knew?
That question became the premise of Scribe, the software company she co-founded with Aaron Podolny in 2019. Its first useful promise was straightforward: do a task on your computer, and turn the steps into a guide someone else can follow. A workaround could acquire an audience without its inventor acquiring a second job as a technical writer.
A career spent asking people to explain
Smith grew up in upstate New York, near the Canadian border. Becoming a startup founder was not a childhood plan. At Princeton, she was still figuring out what she wanted to do. Consulting appealed because the people recruiting her seemed interested in the world and willing to tackle difficult questions.
Her education eventually included an MBA from Harvard Business School. Her working life took her through McKinsey, Coatue Management, and Greylock Partners. That is a résumé with plenty of respectable destinations. The interesting part is the question that kept following her between them.
At Greylock, she spoke with more than 1,200 executives about enterprise technology and the problems they wanted solved. A conversation about software purchasing can easily become a conversation about whatever has the most persuasive demonstration. Smith was listening for the work underneath it: what people needed to accomplish, where they struggled, and what technology had left untouched.
Years after those operations-center visits, the difficulty of capturing practical expertise was still there. Companies could buy more software and build more elaborate repositories. Someone still had to explain how the work happened. Then someone had to write it down. The explanation could be excellent and the system for distributing it painfully ordinary.
Smith has called herself an accidental entrepreneur. “To me I fell in love with a problem,” she said. The description fits a career in which repeated exposure mattered more than a sudden flash. She had seen the gap from the employee’s desk and heard about it from executives. Eventually, leaving it for somebody else to solve became less appealing than starting a company.
Six people, and a useful change of mind
Scribe’s early direction involved computer-vision-based automation. Customers helped rearrange the priorities. The part that let them understand how people worked was valuable in its own right. Before deciding what a machine should take over, a business needed a clearer account of what was happening.
Smith and Podolny kept the early company small. In 2021, she described a team that had consisted of the two founders and four engineers. Her phrase for the approach was “painfully lean.” Scarcity forced decisions about which requests mattered enough to build. Every addition had to compete with something else the team could do.
The first version released through Product Hunt in 2020 saved its output locally on a desktop. It was not even internet-enabled, Smith later recalled. For software intended to help people share knowledge, that was a rather pointed limitation. The founders released it free to find out whether people cared enough to use it anyway.
They did more than use it. Feedback arrived in detail. In her 2024 conversation with investor Sandhya Hegde, Smith remembered strangers sending pages of requests. Some of the requests were basic: let me edit this; let me share it as a web link. Frustration was evidence of attachment. People were describing the distance between a crude tool and something they wanted in their working lives.
Smith also recalled committing to the product direction despite missing numerical goals the team had set. She remembered the deadline as August 31. The specificity matters: a founder’s conviction can sound like an abstract virtue until it has a date and an imperfect set of results attached to it. Here, the intensity of the response helped her make the decision.
The office expert gets a little breathing room
Scribe Capture made the original interaction easier to distribute. A user could record a process and receive written instructions with screenshots, then edit and share the guide. Its natural path through an organization was built into the task: the person making the instructions usually had someone in mind who needed them.
That gives the product a social dimension beyond documentation. In a 2022 conversation with Allison Pickens, the question was how experienced people could help newcomers without losing every spare minute to requests. Being the colleague who knows the answer is flattering. Being asked the same question repeatedly is a less glamorous form of recognition.
A reusable guide changes the arrangement. The expert can offer help once and make it available again. A newcomer can consult the steps without arranging another interruption. There will still be exceptions, judgment calls, and reasons to speak to a human. But the routine explanation has somewhere to live.
Smith’s language about expertise is deliberately inclusive. In an interview about peer learning, she put it simply: “I truly believe everyone is an expert.” The claim gives the office workaround a different status. A person does not need a distinguished title to know something useful about a process. Often, repeated contact with the task is the qualification.
Her ambition places value on those people while trying to loosen the dependence on them. A company can celebrate its go-to colleague and still give that colleague an afternoon with fewer interruptions. In Smith’s account of Scribe, sharing what someone knows should also help that person get recognized for knowing it.
“I truly believe everyone is an expert.”Jennifer Smith
- 01Do the workAn experienced colleague completes a task.
- 02Capture the stepsScribe creates a guide with screenshots.
- 03Pass it onReview, edit, and share the instructions.
Learning to be seen
The founder who wanted to make expertise visible had her own adjustment to make. In a reflection on 2023, Smith said she had been fiercely camera shy before starting Scribe. She used to hide her face in photographs. A friend even hired a photo coach to help with her wedding pictures.
By the time she wrote that post, she was describing a day on set with Podolny as fun. She had practiced a part of the job she once avoided and discovered that her response could change. It is a modest anecdote, and a useful one. The public-facing CEO was something she learned to become.
Her writing also wanders beyond software. In her reflections on 2022, she described a fictional woman who trades being remembered for immortality, then discovers that her ideas can still be remembered. Smith was interested in the loophole: an idea could survive even when its author vanished from memory. It is easy to see why that premise would appeal to someone building a business around knowledge that outlasts a conversation.
There are less literary signs of attachment, too. She has written about keeping a handwritten customer thank-you note on the office wall. Software companies accumulate dashboards. A note offers another kind of evidence: a person took the trouble to say that something had helped. The habit sits comfortably beside her repeated emphasis on listening.
The money makes a headline. The work continues.
In February 2024, Scribe announced a $25 million Series B. Smith wrote that she had debated making the announcement because a funding figure could consume the story. She wanted people to remember the purpose of the company and the people it was building for. A financing announcement is an awkward place to object to financing announcements, which gives her hesitation some charm.
The figures grew. In November 2025, Scribe announced a $75 million Series C led by StepStone at a $1.3 billion valuation. The company said it had documented more than 10 million workflows across 40,000 software applications. Those were company-reported measures of scale, accumulated from the everyday actions that first interested Smith.
Scribe Optimize extended the ambition from recording a process to analyzing how it could improve. Smith explained the sequence with a comparison to self-driving cars mapping roads before carrying passengers. Her point was about order: establish what happens before deciding how to change it. The old consulting question had acquired a new technical setting.
In May 2026, Scribe announced that annual recurring revenue had passed $100 million. Smith’s public update paired that milestone with a memory of the first customer deal, worth $7,000. She also distinguished the scale of the user community from the paying enterprise business: more than six million users across over 150 countries, and 90,000 enterprise customers. Those are different measures, and both belong in the picture.
The button she wanted to remove
On October 6, 2026, Smith introduced Autocapture. Scribe described it as a way to discover work as it happens and draft documentation without a person remembering to press start and stop. Teams choose the business applications and participants involved; people review, edit, and publish what is useful. The company reported more than seven million Capture users.
The latest product returns to the inconvenience in the original story. Even when an employee has found a better method, turning that discovery into an explanation requires effort. Remove some of that effort, and more of the practical knowledge might become available to the next person, or to an AI agent that needs the same context.
Smith’s aspiration now includes helping businesses make their own accumulated expertise usable by AI. It is a continuation of the question she carried out of consulting. The binder, the slides, the guide, and the agent all need an accurate account of the work. Somewhere, a colleague has already figured out a better way. Smith has spent years trying to make that discovery easier to pass on.