Jannik Reinhard

A budget alert, a Copilot Credit allocation and a usage stop are different controls.

An answer, a finished task, or work that keeps moving while you are away?

An enterprise agent control plane needs five layers: identity, catalog, policy, telemetry and lifecycle.

Build a policy assistant with Microsoft Foundry and Python.

MCP tool production needs more than a working connection.

Copilot helps with everyday work.

An MCP server is easy to connect, but an enterprise still needs identity, limits, monitoring and a clear tool boundary.

Local AI agents can work with files, tools and user permissions directly on Windows devices.

Work IQ gives AI assistants governed access to Microsoft 365 context.

Debug a real Foundry agent that skipped its tool.

This is my complete desk setup for 2026, with the monitor, keyboard, mouse, audio and lighting I use every day.

In this blog post I explain what Fabric data agents are and how you build one.

Explore a Microsoft Foundry landing zone in the portal: shared services, identity, networking, model policies and the tests needed before a wider rollout.

My post about CLI tools vs MCP became my most successful post so far, and one follow-up question came up again and again: where do Agent Skills fit in?

In this blog post I show you how to run Claude Code inside your GitHub pipelines.

Watch my Microsoft Foundry guardrails walkthrough: review filters, assignments and compliance options, then distinguish configuration from tested protection.

In this blog post I explain how multi-agent orchestration works in Microsoft Foundry today.

I show how my agentic operating system connects Notion, Outlook, Codex, OneDrive, GitHub, reusable skills and schedules with human review.

In this blog post I take a look at the Security Copilot agents in the Intune admin center.

A beginner-friendly Microsoft Foundry model deployment guide: choose a model, name the deployment and test the right fit before an app uses it.

In this blog post I take a deep look at Foundry IQ, the knowledge layer in Microsoft Foundry.

A beginner-friendly Microsoft Foundry setup: create a project, understand models and deployments, then run one focused first test.

Build your first agent in the Microsoft Foundry Playground.

Since June 2026, Intune enforces Multi Admin Approval on app-authenticated Microsoft Graph API calls.

In this blog post I explain how observability works for AI agents in Microsoft Foundry: zero-code tracing with OpenTelemetry, the monitoring dashboard, continuous evaluation on live traffic and alerts.

Talk to live screen information with a Microsoft Foundry Realtime voice agent.

In this blog post I explain the new auto-update feature of Enterprise App Management, which is generally available since the June 2026 service release.

Migrate from the deprecated Azure AI Inference SDK to OpenAI v1 before 26 August 2026.

In this blog post I take a deep look at the model router in Microsoft Foundry and the model catalog behind it.

Microsoft Foundry June 2026 added Claude, agent publishing, Toolboxes, Routines and Memory.

In this blog post I explain how I secure Microsoft Foundry end to end: private networking for agents, Entra Agent ID and RBAC, customer-managed keys and guardrails.

Microsoft now offers two big platforms for building AI agents: Microsoft Foundry and Copilot Studio.

This is the first episode of my new unboxing series with Admin By Request.

In this blog post I show how to evaluate AI agents in Microsoft Foundry.

In this blog post I explain what Agent Skills are and why they are becoming an open standard across AI tools.

In this blog post I show you how to build your first AI agent in Microsoft Foundry.

AI agents like Claude Code and Copilot CLI run with your privileges.

Service Endpoints vs Private Endpoints, explained simply — where VNet Integration fits, and when to use an NSG, Azure Firewall or WAF.

In this blog post I explain how Microsoft Intune really works under the hood on Windows — including the new MMP-C and Declared Configuration channel — where every log lives, the tools I use, and a clear method to fix the errors you hit most.

A plain, hands-on look at the latest AI models in June 2026 — OpenAI, Anthropic, Google, Microsoft MAI, Mistral and the Chinese open-weight models.

How to run AI models fully offline on a Mac with Microsoft Foundry Local — no Azure, no API key, no internet.

A fact-based look at Intune Advanced Analytics from someone who ran AIOps in a large enterprise — what it does, how it works, and how it really compares.

Build 2026 was not a feature update — it was Microsoft declaring the agentic era a platform.

A hands-on review of the Oakywood Standing Desk Pro — and why it actually improved my daily setup and productivity.

Skills, MCP servers, CLI tools, Computer Use, function-calling, declarative agents — the AI tooling surface has multiplied.

Microsoft 365 Agents — the official term for the agentic platform inside the Microsoft 365 stack — moves AI from chatbox to autonomous worker.

Discover an open-source Microsoft Intune tool using Azure OpenAI to generate policy descriptions and detect conflicts across your tenant.

Why CLI tools often outperform MCP for AI agents: less context overhead, learned commands, and better composability.

Azure AI Content Safety explained: configure harm filters, Prompt Shields, groundedness, protected material and production guardrails in Microsoft Foundry.

My daily productivity toolkit: Akiflow, Raindrop, Claude, Codex, Bitwarden and more tools that keep work simple and focused.