Sovereign Enterprise AI Platform
The sovereign AI platform for enterprises
ai:u is a sovereign enterprise AI platform that securely connects your internal company data with state-of-the-art AI models and assists employees in their daily work through intelligent chats, agents, and browser extensions. As part of our infra:u cloud stack, the platform runs fully GDPR-compliant and without vendor lock-in on our hardware in German data centers.
To take this independence further, ai:u builds the bridge. We create data sovereignty through a holistic approach that permanently resolves dependencies on proprietary providers. The entire solution runs on our infra:u Sovereign Cloud Stack (SCS) and processes all data on our hardware in German data centers. This guarantees technological control and sovereignty from the infrastructure layer all the way up to the application layer.
With the large frontier models (such as those from OpenAI or Claude) as well as the cloud services of the hyperscalers, it is often not conclusively clear how your business secrets are being processed. Using them brings massive risks for digital sovereignty:
Alongside the dreaded vendor lock-in and unresolved use of training data, the US CLOUD Act looms over European companies. It allows US authorities to access data held by American IT corporations, even if the servers are located in Europe. There is also the geopolitical "US kill-switch risk": the danger that business-critical IT services are abruptly shut down under foreign government pressure.
ai:u reverses this principle. Your data is processed exclusively in isolated models on our infra:u cloud stack in German data centers: no US provider access, no kill-switch, clear tenant separation.
The desire for in-house open-source AI often fails in practice: the models require massive GPU resources (VRAM), procuring AI servers is extremely expensive, and building it out requires specialized know-how.
ai:u solves this problem. You benefit from the security and control of our operating model without having to run the hardware yourself. We deliver the most powerful open-source models to you turnkey from our cloud, billed by consumption or, on request, via dedicated GPU resources for your company.
ai:u consists of several modules that work seamlessly together to bring AI from experimentation into productive daily work:
The AI:U API always provides up-to-date open-weight models (such as Llama 3, Mistral, DeepSeek, Qwen, and GLM). Alongside text models, AI services for transcription, speech output, image/video processing, embeddings, and reranking are also available. All processing takes place in our cloud stack in German data centers. The associated portal offers you user management, OIDC integration, guardrails, financial limits, and complete reporting.
AI:U Chat is the secure, browser-based chat client for your employees. Here you can define your own skills and agents and share them with your team.
With the AI:U Browser Overlay (browser add-on), you also bring AI exactly where the work happens: results of AI tasks are displayed directly within your existing web applications. For example, support staff see a suggested solution live within the running ticketing system.
Hand off recurring, time-intensive tasks such as reports, evaluations, or error analyses on IT systems to autonomous agents. These access your connected company data and complete tasks in the background.
Artificial intelligence unfolds its real value only once it knows the context of your company. The AI:U Hub is the data bridge between our AI models and your existing systems.
We integrate AI into your existing system landscape. For example, we connect systems like SAP (ERP) or DATEV (accounting), but the principle applies to any type of business software: your CRM system, complex ticketing systems, internal company wikis, document management systems (DMS), or highly specialized industry software.
Through permission concepts, you strictly define which data is accessible to which users and AI agents. Processing of this sensitive company data never leaves the hardware we control in German data centers.
- EU AI Act & GDPR-ready: Certified operation (ISO 27001) to the highest European standards.
- No provider-side processing: We do not process your data for our own purposes. There is no training on your prompts or documents.
- Tenant separation: Every customer is strictly isolated at the database and infrastructure level.
- No kill-switch: Operations are run in German hands. No one can cut off your access for geopolitical reasons.
- Open standards instead of a black box: We rely on open models and interfaces. You do not become dependent on a single proprietary algorithm.
FAQ
Frequently asked questions
Yes. Since all data is processed exclusively on hardware we control in German data centers and no data flows to third-party providers (as with US hyperscalers), ai:u meets the highest compliance requirements, including the strict provisions of the EU AI Act and the GDPR.
No. We take care of the resource-intensive hosting of the models. You draw on AI capacity flexibly as a service.
Yes. The AI:U Hub is designed not only to connect standard software like SAP or DATEV, but to communicate via common interfaces with almost any knowledge base, DMS, CRM, or ticketing system.
We provide leading open-weight models such as Llama 3, Mistral, DeepSeek, Qwen, or GLM and keep them continuously up to date. You choose the model best suited to your specific use case (e.g. coding, text generation, or RAG).
Billing is 100% transparent and, by default, pay-as-you-go based on tokens consumed (in, out, and cached). For very compute-intensive or dedicated workloads, we agree on an individual pricing model based on dedicated GPU systems.
Do you have more questions? Contact us