Profile · Builders & Makers
DavidSiegel
He builds software for the people who never signed up to build software.
David Siegel · @dvdsgl · San Francisco
He builds software for the people who never signed up to build software.
Ask most founders what they sell and you get a category. Ask David Siegel and you get a verb: glide. Open a spreadsheet, add a few rows, and the thing on the other side is a working app - sortable, searchable, shareable, with a login screen and a layout someone bothered to make beautiful. That is the entire pitch of Glide, the company he co-founded in 2018 and runs as CEO. No download of an IDE. No semicolons. No waiting six months and $400,000 for a dev shop to ship something the team stops using by quarter two.
Glide's bet is older than Glide. Spreadsheets are, by a wide margin, the most widely used programming model ever invented - hundreds of millions of people manipulate data in cells without ever calling it programming. Smartphones are the most successful computer ever shipped. Siegel and his co-founders looked at those two facts sitting next to each other and asked the obvious question almost nobody had built around: what if the spreadsheet you already maintain simply became the app you wished you had?
The answer is now the company's center of gravity. Glide reads your data, infers structure, and renders a polished interface that syncs both ways - change a cell, the app updates; tap in the app, the cell updates. More recently the product has moved past templates entirely. Siegel describes Glide as building the first AI app generator: you describe what you need, and software appears. By his own count, the company is approaching ten thousand apps connected to its AI services, built by people who, in his words, are "really curious" about "a new capability that no one's wielded before."
Glide is not chasing the next consumer hit. Siegel's advice to builders is almost contrarian in a venture world addicted to virality: stop trying to build the next TikTok. The apps that matter on Glide are inventory trackers, field-service checklists, member directories, internal dashboards - the quiet machinery that keeps an organization running and that traditional software economics never priced low enough to serve. That is the design desert he keeps returning to. His own framing for a career is finding neglected territory, the places everyone agrees are too small or too boring, and turning them into what he calls design oases.
It is a useful lens for reading everything he has made. The pattern is consistent: take a capability locked behind expertise, and hand it to the person who actually has the problem.
Spreadsheets are the most successful programming model of all time, and smartphones are the most successful computers of all time. Can these two forces be combined into something very valuable?
- David Siegel, on the idea behind GlideBefore he built tools that erase the need for developers, Siegel spent years building tools for developers - which is the best possible training for the job. He interned at Google in the mid-2000s, working on Google Talk. He prototyped at Idealab. Then came the stretch that shaped him most: as a user experience designer for Ubuntu at Canonical, he ran a project to fix the small, papercut-sized annoyances that pile up in an operating system until using it feels like a thousand tiny wounds.
He did more than fix them. He named them. "Papercut" as a term of art for a minor, overlooked bug entered the open-source vocabulary through his Papercuts project, which closed more than 250 of them in Ubuntu. His line about it has the dry honesty of someone who has triaged a long backlog: "Everyone thinks their bug is a paper cut." The discipline underneath the joke - taking small problems seriously because users feel them every day - is the same instinct that later made Glide's apps feel finished instead of generated.
From there he led design at Xamarin, the cross-platform mobile development company, from 2012 to 2016. When Microsoft acquired Xamarin for a reported $500 million, Siegel became head of design and developer services in the Microsoft Developer Division. He had a front-row seat to exactly how much the world's largest companies wanted mobile apps, and exactly how brutal it was to make them. "We saw how desperate some of the world's largest companies were to have a mobile strategy," he has said, "and also how painful and expensive it is to develop mobile apps." Glide is the answer he wrote to that observation.
Tucked inside his resume is a project that quietly touched an enormous number of developers: quicktype, a tool that generates types and converters from JSON, Schema, and GraphQL. It has collected well over 13,000 stars on GitHub and, by his own description, was the most prolific code generator in the world before GitHub Copilot arrived. It is the sort of thing a person makes when they cannot stop building tools even on the weekend. His earlier GNOME Do, an application launcher, came from the same compulsion. The through-line from a Linux launcher in the late 2000s to an AI app generator in the 2020s is one person repeatedly asking why the useful thing is so hard to reach, and then closing the distance.
"Giving people enjoyable experiences through technology is my life's work."
"We're approaching 10,000 apps connected to our AI services. It's a new capability that no one's wielded before."
"Everyone thinks their bug is a paper cut."
"We saw how desperate the world's largest companies were for a mobile strategy - and how painful it is to build one."
His handle, @dvdsgl, is just David Siegel with the vowels removed - and it doubles as his personal domain, dvdsgl.co.
Before picking spreadsheets as Glide's foundation, the team studied close to 100 different no-code tools.
His personal site mixes essays on Aristotle, Descartes, and Plato with technology criticism - philosophy was half his degree.
quicktype, a side project, became one of the world's most-used code generators before Copilot existed.
His whole career rhymes: take a capability locked behind expertise, and hand it to the person who actually has the problem.