Allbirds has transitioned from its shoe business to an AI infrastructure startup named Smartbird, led by new CEO Nadia Carlsten, who faces the challenge of building a team from scratch to serve a niche market focused on data sovereignty and controlled AI compute deployments.
- Allbirds sold shoe business and raised $100M for AI pivot
- New CEO Nadia Carlsten building team and strategy for Smartbird
- Targeting niche AI compute market needing data control over scale
What happened
Allbirds completed its pivot from footwear by selling its shoe business for $43 million and renaming itself Smartbird to focus entirely on AI infrastructure. The company raised an additional $100 million from the stock market to fund this new direction. Nadia Carlsten, former AWS executive and European compute leader, joined as CEO to lead this transformation starting June 2026.
Smartbird currently operates without a team but is actively recruiting leadership and plans to establish a physical office in Amsterdam. The company aims to build specialized AI compute clusters catering to customers who require direct control over infrastructure, particularly those sensitive to data sovereignty and compliance issues.
Why it matters
Smartbird’s focus on controlled AI infrastructure addresses a niche market distinct from hyperscale public cloud providers. Many organizations in sectors like pharmaceuticals, energy, finance, and public institutions require bespoke compute environments because of regulatory or business model demands, making them underserved by commodity cloud services.
This business model differs from the scale-driven economics of major clouds competing on price and chip utilization. Instead, Smartbird prioritizes agility, security, and data control for clients who value those features more than low-cost compute. As AI adoption grows, the demand for highly customized infrastructure solutions could become a sustainable market segment.
What to watch next
The company expects to deploy its first compute clusters for paying customers by the end of 2026. Key milestones will include successfully assembling a leadership team, establishing operational infrastructure, and onboarding initial clients, which will validate Smartbird’s business model in this early-stage market.
Competition will come from established players offering managed AI compute services like Hewlett Packard and Equinix, as well as internal projects within target companies. Observers should monitor Smartbird’s ability to carve out a defensible niche and whether it can grow without chasing massive scale or price wars typical among cloud providers.