In a field trial last year, a drone equipped with a small AI model identified and struck a target without a radio link back to a command center [RuntimeWire, September 2026]. The demonstration, supported by NATO's DIANA accelerator, was a stark validation of a specific technical philosophy. For some of the world's most sensitive data, the cloud is not an option. The future of AI in defense, critical infrastructure, and regulated industries may depend on never moving that data at all.
This is the core premise of Scaleout Systems, a Swedish deep tech company spun out from Uppsala University in 2018. Its platform enables a technique called federated learning, where a shared machine learning model is trained across hundreds or thousands of distributed devices,drones, vehicles, hospital servers,while the raw data never leaves its original silo [Uppsala University, October 2021]. In an era defined by both AI breakthroughs and escalating data sovereignty concerns, Scaleout is betting that the infrastructure to make this work at scale is not just a research project, but a commercial necessity.
From Academic Framework to Defense Contracts
The company's origins are deeply academic. Its founding team, including CEO Andreas Hellander and CTO Salman Toor, are associate professors in scientific computing at Uppsala University [Scaleout Systems, Unknown]. Their research into large-scale distributed systems led to the creation of FEDn, an open-source federated learning framework designed for both research flexibility and industrial-grade deployment [Scalable federated machine learning with FEDn, Unknown].
Scaleout was formed to productize this work, initially targeting industries like automotive and healthcare where data privacy regulations create silos. The strategic pivot toward defense and cloud-edge AI came later, accelerated by geopolitical shifts and a clear need for resilient, offline-capable AI. The company's participation in the NATO DIANA accelerator, which came with €100,000 in non-dilutive funding, provided both capital and a crucial stamp of credibility for this market [RuntimeWire, September 2026]. Today, Scaleout lists defense giants like Saab and BAE Systems Bofors alongside industrial partners such as Scania and BMW among its collaborators [LinkedIn, Unknown].
The Sovereign Edge AI Stack
Scaleout's offering, now branded as a sovereign edge AI platform, consists of two main layers. The first is the orchestration layer, evolved from FEDn, which manages the federated learning process across a network of clients. The second is the edge runtime, which deploys and executes the trained models directly on devices in disconnected environments.
The technical challenges here are profound. Federated learning must handle unreliable networks, heterogeneous hardware, and stringent security requirements. Scaleout's architecture is designed to be resilient to these conditions, allowing a model to improve iteratively as it encounters new data on the edge, even when a central server is only intermittently available [Scaleout Systems: Shaping the Future of Federated Learning, Unknown]. For a customer, this could mean training a computer vision model for drone reconnaissance across an entire fleet, with each sortie contributing to a smarter, shared model without any sensitive imagery ever being transmitted.
The Funding and the Founding Team
Scaleout has raised a total of approximately $5.5 million to date, according to PitchBook data [PitchBook, 2026]. Its most recent disclosed round was SEK 35 million (about $3.3 million) in February 2025, led by Fairpoint Capital with continued participation from early backer Navigare Ventures [AI training platform Scaleout secures SEK 35m in a new funding round, February 2025]. The round valued the company at SEK 132 million pre-money.
The team blends academic rigor with enterprise and defense experience. Beyond the professor-founders, the company has recruited engineers with backgrounds at Google, Saab, Oracle, and Boston Consulting Group [Scaleout Systems, Unknown]. This mix is deliberate, aiming to bridge the gap between cutting-edge research and the hardened requirements of national security and industrial contracts.
| Role | Name | Background |
|---|---|---|
| CEO & Co-Founder | Andreas Hellander | PhD in scientific computing, Associate Professor at Uppsala University [Scaleout Systems, Unknown] |
| CTO & Co-Founder | Salman Toor | Associate Professor in scientific computing, former senior cloud architect at Sweden’s national computing infrastructure [Scaleout Systems, Unknown] |
| COO & Co-Founder | Daniel Zakrisson | Not specified in sources |
| CMO & Co-Founder | Jens Frid | Not specified in sources |
| CCO & Co-Founder | Ebba Kraemer | Not specified in sources |
The Risks of a Nascent Category
For all its promise, federated learning remains a complex solution selling into long, conservative sales cycles. Scaleout's ambitions face several credible headwinds.
- Market education. The concept is still unfamiliar to many potential enterprise buyers, who may default to traditional, centralized AI cloud services despite their data governance shortcomings.
- Technical complexity. Implementing a federated learning system requires significant upfront integration work and ML ops expertise, which can be a barrier even for motivated organizations.
- Competitive landscape. While no direct competitors are named in Scaleout's sources, the field is attracting attention. Major cloud providers are developing their own federated learning tools, and other startups are exploring adjacent approaches to privacy-preserving AI.
The company's answer to these challenges is its focus on specific, high-stakes verticals where the value proposition is non-negotiable. In defense and aerospace, the ability to operate in disconnected, contested environments is a requirement, not a nice-to-have. In regulated healthcare or automotive testing, data simply cannot leave a geographic jurisdiction or a company's firewall. By anchoring its early business in these domains, Scaleout aims to prove the model before expanding.
What to Watch in the Next Twelve Months
The coming year will be critical for Scaleout as it deploys its new capital. Key milestones will likely include announcing its first major production contracts with defense partners, expanding its partner integrations,it already has a collaboration with Oracle [Oracle and Scaleout bring Federated Learning to the Tactical Edge, Unknown],and further hardening its platform for deployment at true tactical scale. Another funding round may be on the horizon to fuel growth if these commercial milestones are hit.
The ultimate test for Scaleout's technology is not in a controlled demo, but in daily operation. For a soldier relying on a drone's autonomous identification, or a manufacturer training a quality-control model across ten global factories without sharing proprietary designs, the standard of care today is often a patchwork of manual processes, isolated models, or risky data transfers. Scaleout is betting that a more intelligent, collaborative, and secure alternative is not only possible but urgently needed. Its success hinges on convincing the world's most cautious institutions to trust a fundamentally new way of building AI.
Sources
- [AI training platform Scaleout secures SEK 35m in a new funding round, February 2025] | https://www.scaleoutsystems.com/post/scaleout-secures-new-investment-round-to-accelerate-cloud-edge-ai
- [Ars Technica, September 2026] Small AI models let drones autonomously identify and attack… | https://arstechnica.com/ai/2026/09/nato-backed-startup-adapts-ai-for-autonomous-drone-recon-and-attack-missions/
- [LinkedIn, Unknown] Scaleout | LinkedIn | https://www.linkedin.com/company/scaleout-systems/
- [PitchBook, 2026] Scaleout Systems funding data | Not available
- [RuntimeWire, September 2026] Scaleout’s edge AI picked and struck a target without a radio… | https://runtimewire.com/article/scaleout-edge-ai-autonomous-drone-targeting-alma
- [Scalable federated machine learning with FEDn, Unknown] Academic paper on FEDn framework | Not available
- [Scaleout Systems, Unknown] Company website | https://www.scaleoutsystems.com/
- [Scaleout Systems: Shaping the Future of Federated Learning, Unknown] Article on company's technology | Not available
- [Uppsala University, October 2021] Tech company founded at Uppsala University newcomer on Swedish 33-list | https://www.uu.se/en/news/2021/2021-10-28-tech-company-founded-at-uppsala-university-newcomer-on-swedish-33-list
- [Oracle and Scaleout bring Federated Learning to the Tactical Edge, Unknown] Partnership announcement | Not available