Setup & Configuration (optional)
Upon request: Complete setup of your AI servers including operating system, GPU drivers, CUDA, Docker, Kubernetes or your preferred orchestration. Installation and configuration of AI frameworks like PyTorch, TensorFlow, Ollama or vLLM according to your requirements.
Proactive Monitoring
24/7 monitoring of all critical system metrics: GPU utilization, temperature, memory, network and application performance. Automated alerts for anomalies so we can step in before operations are affected. Grafana dashboards with insight into your infrastructure are available as an option.
Security & Updates
Regular security updates for operating system, GPU drivers, and all installed components. Automated patch management processes with rollback capabilities. Firewall configuration, SSH hardening, and proactive vulnerability scans.
Backup & Disaster Recovery (Optional)
Automated backups of your configurations, models, and data available (optional). Secure storage in geographically separated data centers. Tested recovery processes with defined Recovery Time Objectives (RTO) and Recovery Point Objectives (RPO).
Support & Troubleshooting
Direct access to experienced DevOps and AI infrastructure experts via email, phone, or ticket system. Response times according to the agreed service level. Support with performance optimization, scaling, and troubleshooting of AI workloads.
Service Level Agreements
We agree response and recovery times individually and put them in writing - graded by how critical your environment is. This includes defined priority levels, traceable incident documentation and a regular report on incidents and maintenance work.
AI server maintenance and operations: have your existing AI infrastructure looked after
Your AI application is running, but nobody on the team keeps drivers, containers and the inference runtime up to date. We take over technical operations, monitor around the clock and document every change.
- Managed AI server & runtime
- Operation of existing AI servers
- 24/7 monitoring
- Documented changes
Overview
When the AI server runs, but nobody maintains it
Many AI applications are built in a project and then run on a server nobody feels responsible for. We take over the technical operation of your AI infrastructure: operating system, GPU drivers, containers and the inference runtime such as vLLM or Ollama.
We start with a technical review of the existing stack. Then responsibilities, service level and change paths are agreed in writing; you keep full control over data and models.
Service scope
What AI server maintenance and operations cover
We take over the ongoing tasks around your AI server infrastructure.
Responsibilities
Who is responsible for what
The exact split is agreed in writing after the technical review. A typical distribution:
| Area | WZ-IT | You |
|---|---|---|
| Operating system, GPU drivers, CUDA | Updates, hardening, monitoring | Approval of maintenance windows |
| Containers and orchestration | Operation, updates, restart | Requirements for new services |
| Inference runtime (vLLM, Ollama) | Maintenance, configuration, fault analysis | Requirements on models and context |
| Models | Deployment and testing after approval | Selection and approval of new versions |
| Application and business logic | Optional: Managed LLMOps | Business responsibility |
Infrastructure
Your hardware, your hosting provider or our data centre
We look after existing AI servers or provide new ones in German data centres.
Your hardware or ours
We operate AI servers on your own hardware or at your hosting provider, after a technical review. New hardware is provided as a Managed GPU Server: with one NVIDIA RTX PRO 4000 Blackwell (24 GB) or one RTX PRO 6000 Blackwell Max-Q (96 GB), or with two to four RTX PRO 6000 for large models.Server Location Germany
Servers provided by WZ-IT are located in German data centres. The respective data centre operator is certified to ISO 27001 and BSI C5 Type 2 - these certifications belong to the data centre operator, not to WZ-IT. Your data stays within German jurisdiction. Redundant power supply, cooling and physical access control are standard in these data centres.High-Performance Network
Direct connection to European internet backbones with low latencies. 1 Gbit/s included, 10 Gbit/s optionally available. DDoS protection and redundant network paths for maximum reliability.Flexible Storage Options
NVMe SSD storage for maximum I/O performance during model loading and data preprocessing. Optional connection to object storage (S3-compatible) for large datasets and model repositories. Automated backup systems with encrypted storage.
Pricing
Price and scope
We state the price in the quote, based on configuration and scope of operations. Single-GPU servers are cancellable monthly; for multi-GPU servers the term is stated in the proposal.
Managed AI server & runtimePrice after technical reviewThe price for operations depends on GPU, model runtime and infrastructure scope. It is based on the technical review of your stack; the result and the price are stated in the proposal, all from one source.
Included in operations
- 24/7 monitoring of host, GPU and runtime, security updates for operating system, drivers and containers, maintenance of the inference runtime, incident handling according to the agreed service level and a documented operating environment. Backups are optional.
Payment
- Bank transfer
No Vendor Lock-in: Export of data and configurations in agreed standard formats. You retain control of your AI models and data; if you switch, we support the agreed handover. Term and conditions are stated in the proposal. Note on pricing: Listed prices are non-binding reference prices and may change. The specific price depends on your individual configuration, term, and scope of services. For a binding quote, please contact us directly. All prices are net and exclude statutory VAT. The offers are addressed to businesses.
Why WZ-IT
Why WZ-IT for operating your AI servers?
German Hosting & GDPR
The offered server locations are in Germany. Data flows created by optional integrations, updates and remote access are documented separately, making the privacy assessment of the specific environment easier.AI & Open Source Expertise
Years of experience with open-source AI stacks: Ollama, vLLM, PyTorch, TensorFlow, CUDA optimization. We know the pitfalls of GPU drivers, model quantization and performance tuning, and we bring that hands-on experience into your project.Personal Support
No anonymous ticket support: You have direct contacts who know your infrastructure and your requirements. Fast decision-making, pragmatic solutions, and true partnership instead of call center mentality. On-site meetings possible if needed.No Lock-in Effects
Root access under the agreed operating model and export in agreed standard formats. We use established technologies and document dependencies so that a later transition remains plannable.Flexible Scaling
Start with one server and grow as needed. Easy expansion with additional GPU nodes, storage, or network capacity. We advise you on optimal sizing strategies and support implementation of auto-scaling concepts.Cost Efficiency
For continuous operation, a fixed monthly price is usually easier to plan with than metered cloud GPU hours. No unexpected costs from storage or traffic fees. What works out for your case is something we look at together in the initial call.
Comparison
Managed vs. Unmanaged Comparison
| Managed Service | Unmanaged Server | |
|---|---|---|
| Setup & Configuration | Fully by us | Self-service |
| Monitoring | 24/7 proactive | Self-implementation required |
| Updates | Automated with testing | Manual required |
| Support | Fast expert support | No support |
| Time Investment | Focus on development | Time for admin tasks |
Frequently Asked Questions
Answers to the most important questions
- At the inference layer vLLM and Ollama, plus Open WebUI, containers with Docker or Kubernetes, GPU drivers and CUDA. Further components such as your own Python services or training frameworks are assessed in the technical review, and the scope is set out in the proposal.
- Operating system, GPU drivers, CUDA, containers and inference runtime are updated in agreed maintenance windows. We check driver and runtime versions for compatibility beforehand, because a new driver can break an existing vLLM version. Every change has a rollback path and is documented. Monitoring watches GPU utilisation, temperature, memory and API availability around the clock.
- We agree administrative rights with you depending on the operating model. Your own software and configurations are possible within the agreed technical framework. Change paths are defined so that managed responsibility for the base system remains traceable.
- After the technical review we take over operations on the agreed date. We agree the schedule in the initial call; it depends on the state of the existing stack and the scope of the takeover.
- For a Managed GPU Server from WZ-IT we organise the replacement with the data centre. On your own hardware or at your hosting provider, we detect the defect through monitoring, narrow it down and coordinate the replacement with you or the provider. If the optional backup service is booked, we restore your data from it. Response and recovery times follow the agreed service level.
- No. We take over operations on your own hardware or at your hosting provider if the technical review allows it. If you need new hardware, we provide it as a Managed GPU Server in a German data centre.
- We charge a fixed monthly price. It depends on the number of GPUs, the runtime in use, the number of services and the agreed service level. The basis is the technical review of the existing stack; price and term are then set out in the proposal. New hardware including operations is available as a Managed GPU Server with published prices.
- AI server maintenance and operations cover the infrastructure: hardware, operating system, drivers, containers and inference runtime. Managed LLMOps looks at the application on top: answer quality, retrieval, prompts, model versions, cost and latency per request. Both can be commissioned separately or together.
Related pages
More AI services from WZ-IT
AI services
- LLM hostingRunning models in Germany - model selection, deployment and operation of the inference stack.
- GPU serversDedicated GPU servers with model, vLLM and Open WebUI in a German data centre.
- AI CubeYour own hardware on site when the data must not leave the building.
- Managed LLMOpsContinuously control traces, quality, retrieval, cost and changes in production AI applications.
- Back to the overview of all AI services
In-depth guides on this topic
Related Tutorials & Guides
- Open WebUI, LibreChat, AnythingLLM & Co. Compared 2026Six open-source platforms are widely used for AI chat in the enterprise: Open WebUI, LibreChat, AnythingLLM, LobeHub, Onyx and PrivateGPT. They solve different tasks but...
- Hetzner GPU Servers Compared: GEX45, GEX63 and GEX131 (October 2026)Hetzner rebuilt its GPU range in autumn 2026: the GEX45 arrived on 1 September, the GEX63 on 6 October, and the GEX44 can no longer...
- IT and Digitalisation Funding in NRW 2026: What Is Open and What Has EndedAnyone planning an IT or digitalisation project with public funding in 2026 will find many guides that still mention programmes that no longer exist. go-digital...
- LiteLLM Vulnerability GHSA-7hp6-4w63-5g45: Check Versions and Harden the GatewayOn 30 September 2026, BerriAI published a security advisory for the LiteLLM proxy: GHSA-7hp6-4w63-5g45. An authenticated user with the internaluser role can escalate to proxyadmin...
- Shadow AI in companies: why a policy alone is not enoughIn many companies, AI is in use before anyone has decided on it. Employees draft emails with a private chat account, have contracts summarised or...
- Local, GDPR-Compliant Speech Recognition: Whisper, Voxtral and Parakeet ComparedSpeech recognition is one of the AI applications with the clearest benefit: dictation becomes text, meetings become minutes, phone calls become searchable notes. At the...
- RAG Knowledge Base for Public Authorities: A Guide for the Public SectorPublic authorities manage large collections that have grown over years: decrees and circulars, procedural instructions, technical reports, expert opinions, statements, minutes. People looking for a...
- RAG and the EU AI Act: when a RAG system is high-risk AIAnyone planning a RAG knowledge base soon runs into the question of whether the system falls under the high-risk rules of the EU AI Act....
- Extract and Validate Documents with AI: Self-Hosted Document ProcessingSince 1 January 2025 every German company has to be able to receive e-invoices. Part of the incoming documents now arrive as structured XML and...
- Self-Hosted Langfuse: LLM Observability and AI Logging for the EU AI ActAnyone bringing AI into production business processes quickly faces an uncomfortable question: what is actually happening in there? Which prompt went to which model, why...
- Bleeding Llama (CVE-2026-7482): Why Self-Hosted AI Isn't Automatically Secure AIThree unauthenticated API calls. No login, no exploit framework, no privilege escalation. Three POST requests to a default port, and the machine's memory is on...
- EU AI Act: High-Risk Obligations Deferred to December 2027 - What Still Starts in August 2026The EU AI Act's high-risk obligations do not start on 2 August 2026. Stand-alone high-risk systems under Annex III now have until 2 December 2027,...
- GPT-OSS 120B on the AI Cube: Run OpenAI's Open-Weight Model LocallyOpenAI released GPT-OSS 120B as an open-weight reasoning model on 5 August 2025. Its native MXFP4 quantisation allows OpenAI to position the model for a...
Reviews & projects
What clients say about working with us
From a knowledge base with an AI assistant and local AI integration to data sovereignty and ongoing operations.
AI Platform
We were looking for a secure yet easy-to-use alternative to the large American providers and came across WZ-IT. WZ-IT advised us on choosing the right models and implemented the integration for us. Beforehand, we were able to try everything out calmly on a test instance. Today WZ-IT runs Open WebUI for us on our own domain, with a convenient Microsoft login and open-source models hosted by a German provider. Our team can now work productively with AI without resorting to external tools or shadow AI. Working together is straightforward, and questions are answered quickly and clearly. With WZ-IT we also have a partner at our side for further AI projects. (translated)
Carl-Frederic KornHead of E-Commerce & Online Marketing, Krüger Dirndl GmbH
Knowledge Management & AI
The knowledge from many years in the solar industry lived in the head of our former managing director, and it still lives in the heads of our employees today. We wanted to secure it, not as a folder full of notes, but in a way that lets us ask questions in daily work and get reliable answers. WZ-IT advised us from the start and weighed different approaches and knowledge management systems against each other with us before we decided. What was built is a custom knowledge base with an AI assistant that brings together conversations, notes, documentation and manuals: from the platform through onboarding our team to support in operation. Conversations are transcribed and facts are extracted; we review and correct everything ourselves before it is stored, because we do not want to trust an AI blindly. The system runs in our own environment, and the data stays with us. What surprised us was how efficient it was: WZ-IT used existing open-source software and extended it specifically for our requirements, instead of selling an expensive custom build from scratch, as is often usual with IT service providers. Timo and Robin understood our actual problem, preserving knowledge during the generational handover, and implemented it step by step: from the proof of concept through setup and validation with our own data to productive operation, explaining everything clearly along the way. Next, we want to extend this to the knowledge of our employees, so that this does not depend on individual people either.
Joshua von AmelnManaging Director, solarwerkstatt berlin GmbH
Read client reviews
International
Selected projects
- LiveRemote access platform for distributed water treatment plantsABCO Water Systems · Australia · Secure Access & OT
- 81%AWS migration with substantially lower operating costsEVA Real Estate · UAE · Cloud exit & infrastructure
- 98%Provision new sites in minutes instead of hoursnextGYM GmbH · Germany · Automation & IoT
WZ-IT designs, develops and operates infrastructure and software for clients in Germany and internationally. We deliver projects remotely and continue supporting them in ongoing operations after go-live.
Enquiry
Define the operation of your AI infrastructure
Name the existing or planned stack and the responsibility WZ-IT should take over.
- Straight with Timo and Robin - no sales team, no pitch
- A professional assessment of your situation
- Concrete next steps for infrastructure, software or AI
The conversation is free and non-binding.
WZ-IT's advice on our Azure migration was technically sound and completely non-binding right from the intro call - we took away a great deal.
Jakob ÖschlbergerInno7 GmbH
WZ-IT's advice on our Azure migration was technically sound and completely non-binding right from the intro call - we took away a great deal.