Derek Feehrer Is Building the Camera Crew Inside Your Screen Recorder

In May 2020, Derek Feehrer walked away from a job he liked at HubSpot with a career ladder in front of him and about sixty dollars in revenue behind him. Sixty. Barely dinner for two in San Francisco, and an impolite answer to anyone asking for financial projections. Yet the number mattered. It had arrived from a thing he built, and from a customer who was neither friend nor family.

The thing was Appointment Scanner, a narrow service that watches for cancellations in the clogged interview calendars for Global Entry, SENTRI, and NEXUS. Feehrer built it on nights and weekends while working as a front-end engineer on HubSpot's growth team. It checked a public schedule so its users did not have to spend their days refreshing one. The idea was modest, practical, and gloriously free of a manifesto.

More important, it answered a private question. Could he identify a real irritation, build the whole solution, and persuade strangers to pay? Once the answer was yes, however softly, the sixty dollars became less a revenue figure than a permission slip.

Feehrer had spent nearly two years at HubSpot. He liked the company and his colleagues, and he could see routes into engineering management. The trouble was breadth. A role at an established software company could sharpen one part of his craft, while the work tugging at him demanded all of it: design, marketing, backend engineering, pricing, support, and the delicate business of putting a new object in front of the public without hiding behind the org chart.

He described it as a persistent itch. The word suits his career better than the usual thunderclap origin myth. Feehrer's products have emerged from repeated, irritating chores. Appointment Scanner removed calendar refreshing. DashDisplay turned a Tesla browser into a monitor. MockRocket attacked the fiddly work of making software look good in motion.

This was not blind appetite for risk. Feehrer laid out the arithmetic in public. He was 24, had savings, and believed that full-time work would tell him faster whether MockRocket had a future than another succession of nights and weekends. May 15 was his last day at HubSpot. Five weeks later, he wrote that he was nearing a major MVP milestone. There were no champagne metrics to report. There was simply more day available for the work.

The irritant behind MockRocket had accumulated over hundreds of hours. Feehrer made promotional videos for his own apps in tools such as Photoshop and After Effects. A convincing phone or laptop mockup could require 3D models, materials, reflections, shadows, specialist plugins, and a patient relationship with rendering. He could get the result he wanted. He resented the pilgrimage.

MockRocket put the task in a web browser. A user could drop a screenshot or video onto a realistic device, animate it, customize the scene, and export at up to 4K. Its rendering used WebGL and WebAssembly in the browser. That technical decision doubled as a product promise: creative software could be immediate, accessible across machines, and private enough that imported designs did not have to travel to a server.

The first public release took about a year and a half. In January 2022, Feehrer brought it to Hacker News with a restrained invitation for questions and feedback. The post earned 75 points and a 30-comment discussion. One commenter praised the rendering fidelity and the way browser-based creative tools widen access. Feehrer, meanwhile, called the launch what it was: the first release.

There is a small piece of temperament in that phrasing. He had worked long enough to know how hard the thing was, but not so long that he mistook shipping for completion. MockRocket solved the visual mockup. Customers wanted the larger movie: projects they could save, multiple clips they could combine, text, music, annotation, recording, and a process that reached all the way from raw interaction to finished demo.

By 2023, the solo-founder arrangement had begun to feel as limiting as the old specialist job. Feehrer's bootstrapped products were supporting him, but managing several of them alone left little bandwidth. Remote work had also made the independence lonelier. Todd Ashley, whom Feehrer had met in Boston roughly five years earlier, had become the obvious person with whom to imagine something larger. They shared an interest in software and media. On a hike in Los Angeles that May, they discussed working together, then agreed to sleep on it.

The next day supplied an alarm clock. Feehrer received an unexpected invitation to interview with Y Combinator. He flew back to Boston, preparing for a ten-minute Zoom call on the plane. The interview went well, with one sharp condition: YC wanted a co-founder, and it wanted one immediately.

Feehrer sent Ashley a message headed “RED ALERT!!!” Ashley was at an appointment. He escaped to his car, joined the call, and committed on the spot to leaving his job. It is the kind of founder anecdote that sounds designed by a screenwriter who has been asked to add stakes. Its quieter lesson is better: the call could move quickly because years of trust and a recent honest conversation had already done the slow work.

Soon Feehrer and Ashley were living in San Francisco and building through YC's Summer 2023 batch. Kite launched as a screen recorder for software marketing videos, but the inheritance from MockRocket was visible: automatic animation, 3D device models, text, and music. The new company did not abandon the hard technical work. It wrapped that work around a more complete customer job.

Today, Kite records Mac and iPhone footage and offers a Windows version in beta. Its editing stack includes cursor smoothing, automatic zooms, 3D device mockups, callouts, blur, text scenes, templates, brand kits, music, and AI voiceovers. Multiple recordings can live inside a single project. At the enterprise end, the company lists project sharing, single sign-on, API-driven videos, and custom integrations. An AI Demo Video Agent, now in early access, aims to turn a rough recording into a polished sequence with animation, text, narration, and music.

The ambition is broad, but the company remains small. YC lists two people. That fact shapes Feehrer's public presence. He posts demos, breaks down timelines, answers questions, invites users into the in-app chat, and explains where Kite differs from simpler screen recorders. When discussing price, he has been frank that the team is experimenting. He is equally frank that Kite is not trying to win a race to the cheapest recorder. The wager is that some teams want to replace a bag of editing tools with one coherent studio.

His demonstrations make the philosophy visible. One recent project combined a screen recording, trimmed and accelerated sections, four zoom blocks, an opening and closing laptop animation, three 3D callouts, text positioned around the recording, and AI voiceover. He published the ingredient list because the work is the pitch. A polished clip may look magical. A useful product must reveal that the magic has handles.

There is also humor in the salesmanship. Feehrer's posts tease founders for dreary screen recordings, poke at the polite patience of their mothers, and compare a basic recorder to a Corolla beside a Ferrari. The jokes are abrasive enough to interrupt the feed, then followed by precise feature distinctions and an invitation to ask for help. Mischief gets attention. Specificity earns the second look.

The line from WPI to Kite is not as strange as it appears. As a student, Feehrer worked on an automated sandwich assembly kiosk with a robotics team. He also helped study how a university in Namibia might introduce online courses, and was recognized as both a University Innovation Fellow and a Grand Challenge Scholar. Different domains, same habit: inspect a cumbersome sequence, understand its moving parts, and build a system that makes the result easier to reach.

Founder stories become less useful when every event is edited into destiny. Feehrer did not leave HubSpot because Kite had revealed itself in a beam of light. Kite did not begin as a neat sketch on a napkin. The path was more instructive: a small paid utility made the leap conceivable; an ambitious creative tool exposed a larger job; customers pulled at its limits; a long relationship became a partnership under comically urgent conditions.

Through it all, the subject remained stubbornly consistent. Software products need to explain themselves. Their builders care how that explanation looks. Traditional video tools can demand time, taste, and specialist technique that a small team does not possess. Feehrer kept returning to the gap because he had lived on both sides of it, first as the person fussing with effects and now as the person trying to package those effects into a few legible decisions.

The stealable lesson is not to quit with sixty dollars. It is to notice what the sixty dollars proved. Small products can be instruments of self-knowledge. They reveal which problems hold your attention, which parts of the work you are willing to learn, and whether anyone else cares. Then, if one annoyance survives long enough, it may deserve a larger machine.

Kite is that machine in progress: part recorder, part editor, part motion studio, and increasingly part automated director. Feehrer is still doing what he set out to do in 2020, only with a co-founder, a broader canvas, and considerably more than sixty dollars of evidence. He wanted every creative part of the job. He found one job large enough to contain them all.