In 2011, a nineteen-year-old named Eric Simons talked his way into a spare corner of AOL’s Palo Alto headquarters. He had no employees, a small incubator check, and no more use for a college degree than he’d had the year before, when he skipped it entirely and left Naperville, Illinois, for Silicon Valley with a childhood friend, Albert Pai. The seed money from the ImagineK12 incubator ran out after a few months. Simons and his small team stayed at AOL’s offices anyway, working out of the space until the arrangement became, effectively, a squat — a story local news covered at the time as a curiosity about a kid too stubborn to go home.
That stubbornness is the throughline of Simons’s career. It is also, more than a decade later, the reason a company that was weeks from shutting down became one of the fastest-growing software products anyone had measured.
Building for customers who disappeared
In 2017, Simons and Pai founded StackBlitz — a cloud development environment that let programmers run a full coding setup inside a browser tab, no local installation required. It found a real audience: millions of developers used it, many of them through free integrations with GitHub and popular JavaScript frameworks. What it didn’t find, for years, was a business model to match that usage.
In 2022, StackBlitz raised a $22 million Series A from Insight Partners. The company spent roughly the next year building enterprise features aimed at large corporate customers who had expressed interest during the fundraising process. By the time those features shipped in 2023, the customers who had seemed excited were largely gone. StackBlitz had a widely used product, a well-known investor, and still no reliable way to turn usage into revenue.
The ninety-day bet
By early 2024, the runway math had turned unforgiving — StackBlitz had roughly eighteen months of cash left and no clear plan to extend it. Simons made the call that most founders spend years trying to avoid: he laid off several members of his team, leaving roughly fifteen people, and set a deadline. The remaining team would spend ninety days building one more product. If it didn’t work, the next board meeting would likely be the company’s last.
That product was Bolt — a tool that let anyone describe a web application in plain language and get a working, deployable app back, built with AI and running entirely in the browser, the same technical territory StackBlitz had spent seven years mastering. Simons and Pai prototyped it in February 2024. The category didn’t have a name yet. Within the year, the industry would be calling it “vibe coding,” and Bolt would be one of the products that defined it.
One tweet, no marketing budget
Bolt.new launched publicly on October 3, 2024. There was no press push, no paid campaign — just a single tweet from Simons showing what the tool could do. It spread from there, across X, Reddit, and YouTube, as people posted their own generated apps.
The revenue curve that followed is one of the steepest ever documented for a software product. Bolt crossed an estimated $20 million in annualized revenue within roughly two months of launch, then reached about $40 million ARR within five months — a pace independent trackers and founder interviews subsequently placed among the fastest of any product in software history, developer AI tool or otherwise. Registered users passed two million within the first couple of months and kept climbing, reaching roughly 5 million by May 2025 and more than 7 million by the end of that year.
What makes the number more striking is what didn’t scale alongside it: headcount. The team that shipped Bolt in ninety days stayed close to that size through the growth spike — reports from founder interviews put the company at fewer than 40 employees, with fewer than 10 in sales or marketing roles, well into 2025.
The investors who said no once already said yes
Three years after StackBlitz’s first venture round failed to translate into a durable business, the company went back to raise again — this time with a very different set of numbers. In January 2025, Bolt’s official account announced the result:
“Today we’re announcing $105.5m in funding to take Bolt to new heights! Our Series B was led by Emergence & GV, with participation from Madrona, The Chainsmokers (Mantis), Conviction, and some of the best operators building in devtools & AI,” the company posted on X.
The round valued the company at a reported $700 million post-money — a figure Forbes later confirmed still held as of August 2025. Total disclosed funding across StackBlitz’s history, including the earlier Series A, reached $135 million.
What the near-shutdown bought
By May 2026, Bolt had moved decisively into enterprise territory: the company announced integration with Microsoft Azure and Microsoft 365, added Microsoft’s marketplace as a procurement channel alongside AWS, and built in the compliance and Teams/Copilot integration that large IT departments require before they’ll approve a purchase. Bolt has said that 75% of the Fortune 500 now use the platform in some form — a claim the company has not published independent verification for, but one consistent with the enterprise push.
None of it happened because StackBlitz found a clever growth hack. It happened because a founder who had already proven, at nineteen, that he’d rather sleep in a borrowed office than fold, made the same bet again at scale: one product, one deadline, and a team small enough that everyone on it had to believe it would work. The eighteen months of runway anxiety, the failed enterprise pivot, the layoffs — none of that shows up in the $700 million headline. It’s the reason the headline exists.
Related founders
- Six Months, $80 Million: How Maor Shlomo Built Base44 Alone — a solo Israeli founder who built and sold a vibe-coded AI platform for $80 million in months, without a co-founder or outside investment.
- Pieter Levels Built a Flight Simulator in Three Hours. It Made $1M ARR in Seventeen Days. — another founder whose fastest product ran on a decade of prior distribution work, not a single lucky break.
Facts about StackBlitz’s 2022 Series A, the 2023–2024 near-shutdown, the layoffs, and the ninety-day deadline before Bolt’s launch are drawn from founder-interview coverage on AgentTalk (agenttalk.substack.com, “We had 90 days to ship or shut down”) and Frederick AI’s founder-story series (frederick.ai/blog/eric-simons-bolt). Eric Simons’s early career — skipping college at 18, moving to Silicon Valley with Albert Pai, the ImagineK12 incubator, and working out of AOL’s Palo Alto office — is drawn from his own essay in Fortune (“15 years after skipping college to launch 3 startups…,” fortune.com, January 2, 2026) and 2012 reporting on the AOL office arrangement by CBS San Francisco and ABC News. Bolt.new’s growth figures (ARR and user milestones) are drawn from Lenny’s Newsletter/Podcast (“Inside Bolt: From near-death to ~$40m ARR in 5 months,” lennysnewsletter.com) and The Peel with Turner Novak (“Zero to $20m ARR in Two Months,” via Spotify/Pocket Casts), corroborated by growthunhinged.com’s product-growth case study. The Series B announcement quote is from Bolt’s official account on X (@boltdotnew, January 21, 2025). The $700 million valuation is per Forbes reporting (August 2025); total funding and Series A disclosure per Sacra’s equity research report on Bolt.new (updated May 2026) and Crunchbase. The May 2026 Microsoft Azure/365 partnership and Fortune 500 usage claim are per Bolt’s own announcement.
