Eighteen months ago, Peter Sarlin sold Silo AI to AMD for $665 million, the largest AI acquisition in European history. He could have stayed. AMD gave him a title and a team. Instead, he left, took the proceeds, and started building something most of the industry thinks is five to ten years away from mattering.

His new company is called Qutwo. It launched this week out of Helsinki, funded entirely by Sarlin's family office PostScriptum, with a team of 30 quantum and AI scientists and one unusually bold bet: that enterprises should start running quantum workloads right now, years before fault-tolerant quantum hardware actually arrives.

That sounds contradictory. It isn't.

The conventional wisdom in enterprise tech goes something like this: quantum computing is fascinating, probably transformative, and definitely not ready. The hardware is noisy. The error rates are brutal. Real commercial applications won't show up until the early 2030s at the soonest.

Sarlin doesn't disagree with any of that. What he disagrees with is the conclusion. "The mistake isn't investing too early," he told TechCrunch. "The mistake is assuming you can catch up later. Quantum expertise takes years to develop. By the time the hardware is ready, the companies that didn't prepare will be three years behind the ones that did."

This is the gap Qutwo is designed to fill. Not quantum hardware. Not quantum algorithms. The layer in between: an orchestration system that lets enterprises route workloads across classical, quantum-inspired, and eventually true quantum processors without needing to understand the physics underneath.

The product is called Qutwo OS. Think of it as a routing layer. A pharmaceutical company wants to simulate molecular interactions. A bank wants to optimize a portfolio with thousands of correlated assets. Today, those jobs run on classical hardware. Qutwo OS handles them the same way, but it's built to shift workloads to quantum processors as they become viable, transparently, without the customer rebuilding anything.

The intermediate step is quantum-inspired computing, algorithms that mimic quantum behavior on classical hardware. They won't outperform a hypothetical perfect quantum computer, but they're available now, and for certain optimization problems they're already faster than brute-force classical approaches.

Sarlin's pitch is that the transition from classical to quantum won't be a switch. It'll be a gradient. And Qutwo OS is designed for the gradient.

This isn't a research project hoping to find a commercial angle. IQM, the quantum hardware company co-founded by Kuan Yen Tan, just announced plans to go public on the NYSE through a $1.8 billion SPAC merger. Sarlin is an investor in IQM through PostScriptum. Nokia's former CEO sitting on the board sends a clear signal about where Qutwo sees its enterprise customer base.

Qutwo's first announced customer is Zalando, the Berlin-based fashion retailer. Together, they're developing what both companies call "lifestyle agents", AI tools that go beyond product search to proactively suggest products and experiences. The quantum angle here isn't about cracking encryption or simulating drug molecules. It's about optimization at scale: taking a customer's preferences, inventory constraints, logistics data, and real-time trends, then finding the best combination faster than classical systems can manage.

Whether that actually requires quantum computing today is debatable. Sarlin would probably say it doesn't, not yet. But training the models and building the infrastructure so the transition happens seamlessly when the hardware matures? That's the whole point.

Qutwo doesn't exist in a vacuum. Finland has been assembling quantum infrastructure for years. IQM, the hardware company, has delivered 15 quantum systems to 13 customers and is heading for the NYSE. SemiQon is building quantum semiconductor chips. VTT Technical Research Centre operates one of Europe's most advanced quantum labs. A report from Maria 01, Helsinki's startup campus, published this week showed Finnish startups tripled their total funding in 2025, rising from EUR 102 million to EUR 337 million.

The country punches absurdly above its weight. Five and a half million people, producing quantum hardware companies heading for billion-dollar public listings, a $665 million AI exit to AMD, and now an orchestration layer startup led by the same founder who delivered that exit. Something is working.

Sarlin has said publicly that Qutwo is "an AI company, not a quantum company." The distinction matters. He's not building processors. He's not competing with IBM or Google on qubit counts. He's betting that the real bottleneck won't be the hardware itself, but the software layer that makes quantum useful for companies that just want their problems solved.

Silo AI took seven years from founding to exit. Qutwo is playing a longer game in a market that barely exists yet. The funding model, self-financed through a family office, means no outside pressure on timelines. No board demanding revenue by Q3. Just a team of 30 scientists and a thesis that the quantum era will reward whoever builds the plumbing first.

Sarlin has earned the right to make that bet. Whether the market will validate it is a question that probably won't have an answer for years. But if you're tracking where Nordic deep tech is heading next, Qutwo just planted a flag that's hard to ignore.