01
Foundry setup
Azure subscription preparation, Foundry resource provisioning, network and RBAC concept, logging and cost management. A setup that's production-ready on day one — not POC architecture you have to rebuild later.
Microsoft Cloud · Microsoft Foundry
As a Microsoft Foundry Partner we translate Microsoft's unified enterprise AI platform — 11,000+ models, Foundry IQ as the agentic RAG layer, 1,400+ connectors — for the DACH mid-market. GDPR- and EU AI Act–compliant setup, without hyperscaler complexity, with full architectural control.
What Microsoft Foundry is
Microsoft Foundry bundles what at many companies today is patchwork across individual Azure components: model hosting, retrieval, connectors, governance, and observability. The promise isn't "more models" — it's less architecture effort per use case.
Foundry vs. Copilot Studio
Foundry is the pro-dev platform, Microsoft Copilot Studio the low-code variant. Both have their place — and they mesh well when the setup is right.
Audience. Pro developers, AI engineers, data scientists. Full control over model choice, data pipelines, architecture, and governance.
Typical use case. A productive customer-service agent that orchestrates with knowledge base, tickets, order history, and external systems — and is integrated at the end into Microsoft Copilot Studio or a custom app.
Audience. Citizen developers, power users, business units. Models are abstracted; focus on workflow and data.
Typical use case. FAQ agent for staff, simple lead-qualification bot, automatic approval agent — all without deep model or infrastructure knowledge.
What we deliver for Microsoft Foundry
Azure subscription preparation, Foundry resource provisioning, network and RBAC concept, logging and cost management. A setup that's production-ready on day one — not POC architecture you have to rebuild later.
With 11,000+ models, choice is the bottleneck, not availability. We match use cases to models — latency, token cost, language capability, multimodality, embedding quality — and benchmark before commitment.
Knowledge bases, SharePoint, Microsoft Dataverse, SQL databases, external APIs. We design the RAG architecture to scale and stay maintainable — with clear separation between raw sources, cached embeddings, and answers.
EU Data Boundary, resident models, tenant isolation, Schrems-II-compliant third-country transfer, audit trail. Plus training duty under Art. 4 EU AI Act — see our AI governance advisory.
When Foundry doesn't fit
Microsoft Foundry is powerful but not a universal hammer. We do not recommend it when:
If one of these conditions applies, we say so in the initial conversation. If a Foundry architecture is still sensible, we plan engineering skill build-up in the customer team — that's part of our Application Care model.
FAQ
Microsoft Foundry (Azure AI Foundry until end of 2025) is Microsoft's unified platform for enterprise AI: models, Foundry IQ as the agentic RAG layer, 1,400+ connectors to SAP, Salesforce, Microsoft Dynamics 365, and MCP servers, plus RBAC, network policies, and observability — all under one Azure resource provider.
Microsoft Copilot Studio is low-code for business units and citizen developers. Microsoft Foundry targets pro-devs, AI engineers, and data scientists who build production-ready agents with full control over model, data pipelines, and architecture. Microsoft positions them as complementary — a 'fusion team' strategy.
Foundry IQ is the agentic RAG (Retrieval Augmented Generation) layer on Azure AI Search within Microsoft Foundry. It connects your enterprise data with Foundry agents, without you having to run your own vector database.
We configure Foundry with EU Data Boundary, resident models, tenant isolation, data-processing agreement (DPA) including Schrems-II-compliant third-country transfer, plus audit trail and roles/permissions concept. The training duty under Art. 4 EU AI Act is part of our AI governance advisory.
Microsoft Foundry runs as an Azure service with consumption-based model-token pricing, supplemented by Foundry IQ search operations. There is no classic per-user license model. In the initial conversation we estimate a realistic order of magnitude based on your planned use cases — from POC to productive scale.
Foundry is Microsoft's strategic successor and bundles Azure OpenAI as a component. If you're starting today, choose Foundry — the roadmap, multi-model choice (including Anthropic Claude, Meta, Mistral, xAI), and governance features are only there. We migrate existing Azure OpenAI setups when needed.
30-min initial conversation · 45-min architecture
In the initial conversation we sort out whether Microsoft Foundry, Microsoft Copilot Studio, or a combination is viable — and what a realistic rollout looks like. For more complex setups, go straight to a 45-minute architecture conversation.
Accompanying services
Engineering projects rarely stand alone — license logic, architecture clarification, quality gates, knowledge transfer, and follow-on operations usually run in parallel. Here are the most common accompanying services we add to Discovery Spikes, sprint fixed-price engagements, or Application Care contracts.
Up front · architecture
Before implementation: tenant structure, data model, security concept, integration mapping. The result is an architecture document any engineering team can pick up — including one other than us.
Read more →
Up front · CSP
Which license bundles for which users, which add-on SKUs are needed, where you are over- or under-licensed. Procured via Microsoft Licensing Partner — with the option to use CSP purely as a control mechanism without margin maximization.
Read more →
During · quality gate
Independent second opinion during a running implementation project — whether we are delivering it or another partner. CMMI-based quality gates, risk reviews, fixed price per gate.
During · adoption
Not the classic two-day workshop that's forgotten after a week — but a dynamic learning program over 4–6 weeks with kickoff training, application phases, and advanced sessions. Training matrix for roles and topics.
Read more →
After · operations
After go-live: a predictable Application Care contract with monthly flat rate, SLA-based. Includes releases, hotfixes, extensions, tenant hardening — and continuous support instead of merely reacting to tickets.
Read more →
After · knowledge
When the original developers are gone, the previous partner is no longer reachable, or the documentation is outdated — reverse engineering of the existing solution with a documented result: code map, data model, customization inventory.
Read more →