Peter Sarlin built Silo AI, sold it to AMD for $665 million in 2024, and chose not to pursue "Europe's OpenAI." Instead, he took the capital and a different bet: an AI company that treats quantum computing as just another compute substrate.

QuTwo — his new AI lab — just raised €25 million at a €325 million ($380M) valuation in a single angel round. The investor list reads like a who's who: Yuri Milner, Xavier Niel, Nico Rosberg, Dieter Schwarz, Niklas Zennström, plus founders from Hugging Space, Legora, Miro, Skype, Supercell, and Wolt.

That's not a typical angel round. That's a conviction bet from people who've backed major tech companies.

What QuTwo Actually Is

QuTwo OS is an orchestration layer. The core idea: enterprise AI workloads don't all need the same compute substrate. Some tasks are best handled by classical chips. Some benefit from quantum-inspired optimization. The right architecture depends on the problem type.

Rather than building one model, QuTwo directs tasks to the appropriate compute layer — classical, quantum-inspired, or hybrid. This is the "compute abstraction" layer for AI: you describe the problem, QuTwo routes it to the right underlying architecture.

The company uses "quantum-inspired" computing — classical chips running algorithms that simulate quantum behavior, without the hardware complexity and instability of actual quantum computers. More reliable, commercially available today, with a path to genuine quantum as hardware matures.

Sarlin's framing: "AI is the North Star that we will continue to aim for. Quantum is just a new type of compute." The AI goal isn't the quantum angle — the quantum angle is how they achieve AI performance advantages on specific problem types.

The Team Behind It

Peter Sarlin co-founded QuTwo with Kaj-Mikael Björk (former Silo co-founder) and Kuan Yen Tan (IQM co-founder). IQM is a European quantum computing company, giving QuTwo direct access to quantum hardware expertise and a clear path to genuine quantum integration as the technology matures.

The team has approximately 50 quantum and AI scientists — a substantial research organization for an early-stage company.

The Revenue and Enterprise Traction

$23 million in committed revenue. Design partnerships with Zalando (major European e-commerce platform). Recent expansion into Sweden.

These aren't just forward-looking metrics. $23M in committed revenue at the angel stage — before the full product launch — is a strong signal of enterprise demand for the compute orchestration approach.

What This Means for AI Builders

The QuTwo story has several implications:

Compute diversity is becoming an enterprise problem: As AI moves from experimentation to production, enterprises are discovering that not all AI workloads belong on the same infrastructure. The orchestration layer — deciding which compute substrate handles which task — is becoming as important as the model itself.

Quantum-inspired is the practical quantum entry point: Full quantum computing is still years from commercial reliability. Quantum-inspired computing — classical algorithms that simulate quantum behavior — delivers some of the advantages today. This is the pragmatic bet: build on what works now, migrate to genuine quantum as hardware matures.

AI abstraction layers follow the same pattern as cloud abstraction: In cloud computing, the abstraction layers (containers, Kubernetes, serverless) captured enormous value by hiding infrastructure complexity. AI orchestration — routing tasks to the right compute substrate — may follow the same pattern.

Former successful founders can raise on vision: Sarlin's track record (Silo AI → AMD for $665M) means he can raise €25M on an angel round at €325M before the full product is public. The founder's reputation and network become a competitive advantage in fundraising.

QuTwo at $380M valuation on €25M is a bet that the next phase of AI infrastructure isn't about which model you use — it's about which compute substrate is right for which problem.