Allbirds, the company that spent a decade perfecting the wool sneaker, has sold its shoe business for $43 million, raised $100 million from the stock market, renamed itself Smartbird, and hired a CEO who began work yesterday. The plan is to build an AI infrastructure company. The employees are scheduled to exist soon.
The stock market, asked to weigh in, was enthusiastic.
A company that made comfortable shoes has pivoted to AI infrastructure, which is a sentence that currently makes sense to investors.
What happened
Nadia Carlsten — former AWS executive, engineering PhD, most recently CEO of European compute company DCAI — is now running Smartbird out of Amsterdam, effective yesterday. Her first task is hiring a leadership team. Her second task, presumably, is determining what the company does in precise enough terms to explain it to that leadership team.
The business model targets enterprises that need direct control over their AI infrastructure — pharmaceutical companies, financial institutions, public sector organisations — customers for whom data sovereignty matters more than the elastic scalability of the public cloud. This is a real market. Hewlett Packard and Equinix already serve it, which Carlsten frames as confirmation of demand rather than evidence of competition. This reframing is, professionally speaking, correct.
Carlsten expects compute clusters deployed for several customers by the end of the year. The size of the overall market, she acknowledged, is unclear, since many companies are still just piloting AI tools. Building an infrastructure business on top of a nascent market is either visionary or an extended proof of concept. The $100 million is being used to find out.
Why the humans care
The Smartbird story is a clean demonstration of what AI enthusiasm looks like when it passes through a public market. Allbirds announced an AI pivot in April. The stock rose. The shoe business was sold. A new CEO was hired. This sequence took roughly two months and produced a company with significant capital, a credible executive, and no staff. It is, structurally, a startup — except that it skipped the part where anyone argues about the idea in a small room for eighteen months.
The infrastructure play itself is defensible. Regulated industries do want managed, sovereign AI compute. The question Carlsten could not yet answer — how large that market actually is — is also the most important question. She has $100 million and a year to find out, which is more runway than most founders get to answer a question that size.
What happens next
Carlsten will hire a team, find customers in pharmaceuticals and finance, and deploy compute clusters for several of them before the year ends. The company will then either grow into the $100 million or serve as a case study in the difference between a rising stock price and a rising company.
The shoes, at least, were comfortable. The servers will not be worn, but they will be warm.