Today, most large-scale language models run in the cloud. This means organizations must rely on external servers, incur high computing costs, and relinquish control of sensitive data.
NobodyWho takes a different approach, focusing on Small Language Models (SLM) that run locally on laptops and smartphones. By keeping AI on your device, your data never leaves your control. This setup allows offline use, improves privacy, and eliminates the need for expensive server infrastructure.
A Copenhagen startup has raised €2 million in pre-seed funding to bring smaller, more efficient AI models directly to users' devices. The funding round was supported by early Nordic investors PSV Tech and The Footprint Firm, with participation from Norsken Evolve.
“While the rest of the world is racing to build ever-growing LLMs based on a bigger-is-better approach, we believe the next leap forward in AI will come from making models smaller, local, and human-centric. The future of AI will not be won with scale, but with decentralized models that everyone can run on their own devices,” said Cecilie Wagner Falkenstrom, co-founder and CEO.
A sustainable approach to AI
Founded by entrepreneur and artist Cecilie Waagner Falkenstrom and co-founder and CTO Asbjørn Olling, NobodyWho's engine enables small language models (SLMs) to run locally on laptops and mobile phones, giving organizations and individuals complete control over their data.
The founder explains: TFN“The challenge has been that most developers want to build with LLM without requiring deep LLM expertise. NobodyWho exists to enable that while supporting data sovereignty, reducing cost and energy usage, and eliminating dependence on centralized cloud infrastructure.”
Device-first architecture means data never has to leave the device, enabling true data sovereignty and privacy by design. Running increasingly large-scale cloud-based LLMs, from training to inference, comes with significant power bills and increased CO₂ emissions.
However, the company's local-first SLM architecture significantly reduces this footprint by shrinking models and placing them closer to where they are used. Early benchmarks show up to 100x lower training footprint and up to 500x lower inference footprint.
The platform allows developers to fine-tune and run AI models directly within their mobile or desktop apps without requiring experience with large-scale language models.
With this approach, NobodyWho provides offline, data-secure, privacy-protected, and sustainable AI for both people and the planet. More than 5,000 developers are currently using the company's open source SLM engine.
Unlike OpenAI's GPT model, Anthropic's Claude, Google's Gemini, and NobodyWho, it competes on privacy by design, energy efficiency, and European AI sovereignty.
What about diversity?
When asked about diversity, the company commented, “We are a small and growing team led by a female founder and CEO. As we scale, we are proactively focused on building a diverse and inclusive team across both technical and commercial roles.”
What's next?
This new funding will be used to scale NobodyWho's technology and expand local on-device AI adoption across Europe. The founders conclude: “We want to make NobodyWho the core infrastructure for local AI and take an alternative path to cloud-based AI by providing an open source, fast, private and cost-effective local LLM. We want to help Europe establish itself as a world leader in on-device AI.”
“Supporting NobodyWho was a no-brainer for us,” said Christel Piron, co-founder and general partner of PSV Tech. “This is a great team building Europe's critical AI infrastructure that is privacy-preserving, energy-efficient, and accessible to developers and businesses around the world. It's exactly the kind of technology that PSV Tech is here to support, and we want to see scale from the Nordics.”
Sofie Käll, CIO of The Footprint Firm, said: “By enabling high-performance SLM to run directly on users' devices, we significantly reduce the carbon footprint of AI inference, while enhancing data privacy, reducing operational costs and strengthening Europe's AI sovereignty. NobodyWho is pioneering the infrastructure that makes these ultra-efficient models truly plug-and-play for developers. This is one of the fastest growing emissions categories. This is an exciting opportunity for innovative climate technology in the industry, and we are excited to support a team that can move the industry towards more responsible AI.”
