AI as a Service from a German Cloud
Managed AI Without Your Own Hardware
Without your own hardware
You don’t buy GPUs, and you don’t tie up capital. The AI runs on the NETWAYS Managed Services infrastructure and is ready to go right away.
German Data Center, ISO 27001
Operations are conducted in data centers that are entirely located in Germany and ISO 27001-certified. GDPR-compliant and within the EU.
OpenAI-compatible API
You can integrate existing tools, IDEs, and plugins without any modifications. Simply replace the base URL and API key.
Costs as needed
You pay for what you use. No hardware investment, no lock-in with expensive licensing models.
24/7 Support
You’ll have a personal point of contact and ongoing support through MyEngineer, rather than dealing with an anonymous ticket system.
RAG-ready
Hosted retrieval models and Open WebUI are integrated, making them an ideal foundation for building your own RAG system.
What Often Causes It to Fail
Cloud-based AI is convenient, but with the choice between expensive in-house hardware and U.S. providers that raise privacy concerns, there’s often no suitable option.
GPUs are expensive and in short supply
Purchasing, operating, and maintaining proprietary AI hardware ties up capital and requires specialized personnel. For many projects, that’s too much work.
U.S. Clouds Are a Sensitive Issue
AI as a Service from hyperscalers is up and running in no time. However, this becomes a problem with sensitive content because the data ends up outside the EU.
Getting started is complicated
Model hosting, quantization, GPU availability: it takes a long time before the first request runs. The expertise required for this must first be developed.
Steps to Setting Up Your Own Managed AI
Four steps—just like with any NETWAYS solution. We’ll find the right component for your specific application and provide personalized support throughout its operation.
Analysis & Concept
We'll clarify your use cases, data protection requirements, and desired models. Together, we'll choose the right NWS-AI module.
→ You start with the right service tier instead of having to procure it yourself at great expense.
Setup & Integration
Set up access in MyNWS, create an API key, and integrate the OpenAI-compatible API with your tools, IDEs, and applications.
→ Existing tools will work without any modifications; just swap out the base URL and key.
Commissioning & Data Integration
Provide Open WebUI as a familiar interface and connect it to internal knowledge via RAG (e.g., using the hosted bge models).
→ Ready for the team to use right away, with answers based on real company data.
Support & Operations
NETWAYS Managed Services manages models, updates, scaling, and GPU monitoring, with end-to-end support and a dedicated contact person.
→ You don't need your own MLOps team or to procure GPUs.
Components of Your AI-as-a-Service Solution
The NWS-AI is tiered. You can start for free or go straight to the customized infrastructure, depending on your needs.
Free Open WebUI
Get started for free: a German, GDPR-compliant chat interface that lets your team try out AI right away.
Benefit: Gain initial experience without any setup or risk.
Managed AI Models
Hosted open-source models, such as GPT OSS or bge for RAG, via an OpenAI-compatible API. You don’t have to worry about GPUs or hosting.
Result: productive AI via API, ready to use right away.
AI Individual
Custom AI infrastructure from Germany. We plan the architecture and hosting and handle day-to-day operations, all tailored exactly to your needs.
Result: A proprietary AI strategy without a dedicated platform team.
GPUs on demand
Dedicated GPU resources for your own AI workloads, available on a scalable basis, instead of expensive hardware in your own rack.
Benefit: Computing power on demand instead of tying up capital.
What You’ll Achieve
Ready to Go Right Away · Data Sovereignty · PredictableCosts
Ready to go right away
Up and running without having to purchase hardware or go through a lengthy setup process. The first request will be sent as soon as the API key is set up.
Data Sovereignty & GDPR
Processing takes place in German data centers certified to ISO 27001. Your data stays in the EU, not with a hyperscaler.
Costs as needed
Usage-Based Instead of Hardware Investment. You start small and scale up as needed, without tying up capital in GPUs.
What does our solution run on?
Proven open-source components, operated by NETWAYS Managed Services on German infrastructure. You don’t have to worry about models or operations.
vLLM
The inference backend behind Managed AI Models: high throughput, efficient GPU utilization, and an OpenAI-compatible API—powered by NETWAYS Web Services.
OpenWebUI
The hosted, ChatGPT-like interface. A free, open-source web UI from Germany that’s ready to use right away and GDPR-compliant.
n8n
Grafana
Provides transparency regarding utilization, throughput, and availability, ensuring that managed operations remain predictable and traceable.
We’ll integrate what you’re already using with
NWS-AI can be seamlessly integrated into your tool ecosystem—thanks to its OpenAI-compatible API, no special customization is required. A selection of modules and interfaces.
Hosted Models
- GPT-OSS-120B
- Llama
- Mistral / Qwen
- bge-m3 (RAG)
- bge-reranker
Coding & IDEs
- VS Code
- JetBrains
- Roo Code / Cline
- Claude Code
Operations & Certification
- ISO 27001
- Data Centers in Germany
- 24/7 MyEngineer
- GPUs on demand
Access & API
- OpenAI-compatible API
- API Key via MyNWS
- MCP
- Open WebUI
Own Data (RAG)
- pgvector / Qdrant
- bge Retrieval Stack
- Nextcloud
- Knowledge Bases
Questions & Answers
Frequently Asked Questions About This Solution
Is there GDPR-compliant AI available in the cloud?
Yes. NWS AI operates its models in data centers that are entirely located in Germany and are ISO 27001-certified. Your requests are processed within the EU, not by a hyperscaler outside the EU. This is how AI can be used in compliance with data protection regulations, even for sensitive data.
What is Managed AI?
Managed AI means that a provider hosts, operates, and maintains the AI models for you, including GPUs, scaling, and availability. You use the models via an API or a ready-made interface without having to worry about infrastructure, quantization, or GPU availability.
Can I use AI without my own GPUs?
Yes, that's exactly what NWS AI is designed for. The Managed AI Models run on NWS's GPU infrastructure. You don't need your own hardware—just an API key from MyNWS—and you can get started right away. If needed, additional dedicated GPUs can be booked.
How much does AI as a Service cost?
Yes, that's exactly what NWS AI is designed for. The Managed AI Models run on NWS's GPU infrastructure. You don't need your own hardware—just an API key from MyNWS—and you can get started right away. If needed, additional dedicated GPUs can be booked.
Is ChatGPT GDPR-compliant?
When used directly, the data typically leaves the EU, which can be a concern depending on the type of data. NWS AI offers an alternative that uses the same OpenAI-compatible format. You can continue to use existing applications, but the processing will remain in Germany.
How do I integrate NWS-AI with my tools?
Using the OpenAI-compatible API: Just enter the base URL and API key—that's it. Many IDEs and plugins, such as VS Code, JetBrains, Roo Code, and Cline, work without any customization. For your own data, you can integrate a retrieval stack using RAG; for actions, n8n is also available.
