We believe in data sovereignty, open-source infrastructure, and European AI leadership. Every layer of our stack reflects these values.
Our GPUs and storage infrastructure are hosted exclusively in Helsinki, Finland through Hetzner, a German company with over 25 years of experience in secure, sustainable hosting.
HetznerState-of-the-art reasoning for entity analysis and knowledge synthesis.
Self-hosted embedding model for semantic search across our authority index.
We exclusively use models from Mistral AI, the leading AI company from France. Our embedding pipeline runs on self-hosted E5-Mistral-7B-Instruct, while Mistral Large 3 powers our inference layer.
By choosing Mistral, we support European AI leadership and maintain full control over our model infrastructure—no dependency on US hyperscalers.
Mistral AI — Paris, FranceWe utilize Weaviate, a high-performance open-source vector database headquartered in Amsterdam. By self-hosting Weaviate, we maintain full sovereignty over our data while leveraging its native support for our Mistral-based embedding pipelines.
Weaviate — Amsterdam, NetherlandsAmsterdam HQBoth the technology and development of the world's largest human authority index happens from our headquarters in Amsterdam. We're building the infrastructure that connects AI systems to verified human expertise.
Our mission is simple: in a world drowning in AI-generated content, we ensure that the voices of real authorities remain discoverable and authoritative.
From the blog
Full data sovereignty. GDPR compliant. 100% European.