Fusionality Raises $3.7M to Build the Software Backbone for Fusion Energy

Lausanne-based Fusionality has raised CHF 3 million (approximately $3.7 million) in pre-Seed funding to develop the software infrastructure needed to operate next-generation fusion reactors. The round was led by Founderful and Playfair.

The startup is tackling one of fusion energy’s most difficult engineering challenges: controlling extremely hot plasma inside a reactor. Fusionality is developing software for plasma measurement, simulation and real-time control, helping fusion companies turn complex sensor data into signals that control heating, fuel and magnetic systems within milliseconds.

From DeepMind Research to a Fusion Startup

Fusionality was founded by Federico Felici and Jonas Buchli, two researchers with deep experience in AI, robotics and fusion science.

Federico Felici, co-founder and CEO, is a control engineer with a PhD in plasma physics. Before launching Fusionality, he worked at the EPFL Swiss Plasma Center on the TCV tokamak and spent around two and a half years at Google DeepMind, where he continued working on AI-based fusion control.

Jonas Buchli, co-founder and CTO, is an electrical engineer from ETH Zürich. At Google DeepMind, he led robotics and reinforcement-learning teams. His work with Felici and their colleagues helped demonstrate how deep reinforcement learning could be used to control the magnets that shape fusion plasma. Their research was published in Nature in 2022.

The founders' work focuses on a fundamental problem in fusion: keeping plasma at temperatures of around 100 million degrees Celsius under control and preventing it from coming into contact with the reactor walls.

Building the Software Layer Fusion Companies Need

Fusionality is developing a multidisciplinary platform that combines plasma measurement, software, control and data engineering.

The need is significant. According to the company, more than 30 private fusion companies and government projects face similar challenges when it comes to turning raw plasma measurements into rapid control signals.

Instead of every fusion company developing its own control infrastructure from scratch, Fusionality wants to provide a reusable software layer that can be adapted with relatively little effort. This could help customers reduce onboarding costs and accelerate the development of fusion systems.

Why Now?

The fusion industry is attracting record levels of investment. Private fusion companies received $4.48 billion during the 12 months ending July 2026, while more than $14 billion has been invested across 56 companies since 2021, according to the Fusion Industry Association.

But much of that capital is going toward individual reactor projects rather than shared infrastructure. As more companies work toward commercial fusion, the need for specialised software and operational tools is becoming increasingly important.

Fusionality is positioning itself as part of that emerging infrastructure layer.

Competition and the DeepMind Factor

The opportunity is not without competition. Next Step Fusion already provides plasma modelling, diagnostics and control software to tokamak developers.

There is also an interesting challenge from the founders' former employer. In October 2025, Google DeepMind partnered with Commonwealth Fusion Systems, open-sourced its TORAX plasma simulator, and applied reinforcement learning to CFS's SPARC reactor. That means some fusion companies may have access to sophisticated AI and simulation expertise without paying for an external provider.

For Fusionality, the challenge will be proving that a dedicated, commercial platform can provide enough value across multiple fusion companies and reactor designs.

Investors Back the Team

Founderful and Playfair led the CHF 3 million pre-Seed round.

Playfair investor Felix Neubeck highlighted the fact that the fusion industry is expanding faster than its supply chain and described operational infrastructure as one of the sector's clearest gaps. He also pointed to the founders' unusual combination of academic expertise and practical experience across different fusion-control technologies.

Paolo Ricci, director of the Swiss Plasma Center at EPFL, also welcomed Fusionality's emergence, noting the importance of turning decades of plasma-control and diagnostics expertise into startups serving the global fusion industry.

What Will Fusionality Do With the Funding?

The new capital will be used to expand Fusionality's team in Lausanne, with plans to hire:

• Control engineers
• Computational physicists
• Plasma measurement specialists
• Software engineers

The company will also use the funding to work toward securing its first paying customers.

The Bigger Picture

Fusion energy has moved from being primarily a long-term scientific ambition toward becoming a rapidly developing commercial industry. But building a working reactor is only part of the challenge. Companies also need the software, control systems and operational infrastructure that allow these extremely complex machines to function reliably.

That's where Fusionality sees its opportunity.

By turning years of DeepMind and EPFL research into reusable commercial software, Felici and Buchli are betting that fusion companies will need specialised digital infrastructure just as much as they need advanced hardware.

If Fusionality succeeds, its technology could become an important part of the software stack supporting the next generation of fusion reactors.