Compliance, sovereignty and governance for the Mittelstand — written for owners and the consultants who serve them, not for an engineering team. Every piece here circles the same idea: agents you can trust — controllable, explainable and accountable. No hype, no jargon: just what to do before the EU AI Act lands on 2 August 2026.
Three principles run through all of it — sovereign by default, open by principle, human by design. If you are weighing the alternatives, the side-by-side comparison shows where they land against a foreign cloud chatbot.
Energy-efficiency advisors, tax consultants and MSPs already own SME trust. Here is how to add a sovereign, compliant AI assistant to your services and earn recurring revenue, without running infrastructure.
Which mini PC should run your local LLM? We compare the best AMD Ryzen AI, Intel Core Ultra and Apple Silicon boxes for on-premises AI — from a €490 starter to a 70B-capable workstation.
A look inside Cloud Shield: how the architecture lets a public-facing AI stay sovereign and DSGVO-compliant, with the trust boundary intact end to end.
The model is a swappable engine, not the product. You don’t connect a model to your customers — you connect it to Wegenty’s governance layer. Local by default, frontier on opt-in.
A 70-billion-parameter model needs roughly 40 GB of memory. Here are the 128 GB unified-memory mini PCs and workstations that run a 70B locally — and how they compare on speed, software and price.
Unified memory or discrete VRAM? CUDA, Metal or ROCm? We compare AMD Ryzen AI, Intel Core Ultra, Apple M-series and NVIDIA for running large language models locally — and why the NPU rarely matters.
Memory is the gatekeeper, not raw speed. Here is how to size RAM, VRAM and quantization to the model you want — a plain-language guide to picking hardware for a local LLM.
Cloud AI bills you per token, forever; a local appliance is a one-time cost you own. We break down the total cost of ownership — and where the line crosses for a typical SME.
What the EU AI Act actually requires of small and mid-sized businesses, what is already in force, and how to turn a compliance deadline into a reason to act — without a legal team.
Everything a small business needs to host AI on its own hardware: why local, what it costs, which box to buy, how to stay DSGVO- and EU AI Act-ready, and how to get it installed without an IT team.
Most AI assistants ship your data to a foreign cloud. Sovereign AI keeps the model and the knowledge under your control. Here is what that means in practice — and why it matters under DSGVO.
Anyone can bolt a chatbot onto a pile of PDFs. The hard, valuable part is the boundary between private knowledge and a public-facing agent — and the loop that makes it improve. That boundary is the product.
The whole core is open source — only the compliance layer is closed. Here is the trust boundary, the architecture, and exactly what you can fork, self-host and audit.
Try the interactive demo, or book a 20-minute walkthrough. No data leaves your building, ever.
