All Intune, AI and Azure Blog Posts
Every article on jannikreinhard.com, newest first. Practical notes, deep dives and field-tested ideas around Microsoft Intune, AI, automation, Azure and endpoint management.
Every company has the same gap: the data sits in lakehouses and semantic models, and the people with questions cannot write SQL or DAX. In this blog post I explain what a Fabric data agent is and how you build one. A Fabric data agent is the conversational analytics piece of Microsoft Fabric — your colleagues ask questions in plain English, and the agent generates and runs the queries against your governed data in OneLake. In February I wrote about why CLI tools are beating MCP for AI agents. It became my most read post so far, and one follow-up question came up again and again: "Okay, and where do Agent Skills fit in?" That is a fair question, because Skills vs MCP is not an either-or decision. They solve different problems, and they work well together. In this blog post I explain the difference, compare the context cost, and give you a simple decision guide for when to use which. Most of us use AI coding agents interactively: you sit in the terminal, type a prompt, and watch the agent work. But a lot of the work in a repository is not interactive at all. Reviewing pull requests, triaging new issues, writing release notes, updating documentation — these are standard tasks that repeat every week. In this blog post I show you how to run Claude Code inside your GitHub pipelines with Claude Code GitHub Actions, so these tasks run automatically while you do something else. One agent alone rarely solves a real business process. As soon as you build something serious, you want a triage agent, a specialist for your knowledge base, maybe one that creates tickets — and something that coordinates them. In this blog post I do a deep dive into multi-agent orchestration in Microsoft Foundry. I explain how the story changed three times within one year, how the Microsoft Agent Framework patterns work, and which pattern I would choose for which scenario. At the end you will know exactly where to build your orchestration today so that it does not get deprecated next year. I already wrote about building your first agent in Microsoft Foundry. This post is the next step: making several agents work together. I do not need another AI chat window. I need an agentic operating system that knows where my work starts, where every result belongs and which actions still require me. That difference sounds small, but it is the reason my daily setup keeps working even as I add more agents, repositories, documents and content channels. My current system connects Notion, Outlook, Codex, reusable skills, OneDrive, GitHub and scheduled automations. Codex is my primary execution layer. Claude Code gives me a second coding path when its strengths fit the task better or when I need to distribute a long coding day across both subscriptions. Notion gives me the working context for the day. Outlook owns communication and time. OneDrive remains the system of record for documents and media, while GitHub owns code. The tools are not interchangeable, and that is the point. This post is not a list of apps I like. It is the operating model I use every day: how work enters the system, how I choose what matters, how an agent gets the right instructions, where the output is stored and where I deliberately keep a human approval step. Microsoft is bringing AI agents into more and more admin portals. In this blog post I explain the Security Copilot agents in Intune: the Policy Configuration Agent, the Change Review Agent, and the new Vulnerability Remediation Agent. I show what each agent actually does, how you can enable them, and what you should watch out for. I also cover an important licensing change: Microsoft 365 E5 and E7 tenants get Security Copilot capacity included at no additional cost. Seeing a model in the Microsoft Foundry catalog is not the same as having a model that an application can use. The bridge is the deployment. Think of it as the saved, named configuration that your app talks to. Your code does not call a vague catalog entry; it calls the deployment name you create. That makes the portal form more important than it looks. It connects a model, a deployment type and a name that people and code will depend on. In this post I explain those three choices in plain English before you put a model behind an agent or an application. The video below shows the practical deployment path in German. Here I add the questions that are easy to miss while the portal is open — and a simple path for a first deployment. Every team that builds AI agents ends up building the same thing: a retrieval pipeline that feeds company knowledge into the agent. And most of these pipelines are built again and again, per agent, per project. Foundry IQ is Microsoft's answer to this problem — a shared knowledge layer in Microsoft Foundry that any number of agents can plug into. In this blog post I explain how Foundry IQ works under the hood, how you create a knowledge base and wire it into an agent, and where the limits are. At the end you can decide if it should replace your custom RAG setup. The first Microsoft Foundry screen can feel like a wall of new words. Workspace, project, model and deployment appear almost at once, even though they do different jobs. If you are new to Foundry, start with this: a project is the concrete use case your team works on, a model is the capability you want to try, and a deployment is the named version that an app can call. The word workspace needs a little more care — I explain that below. You can create all three quickly. The useful part is making the first setup easy to understand when you return to it next week. In this post I show the small, practical structure I use before an experiment turns into an application dependency. The video below shows the practical click path in German. This article adds a beginner-friendly explanation around it, so you know what each choice means before you move on. In June 2026 (week of June 22), Microsoft changed something that can break your automation overnight. Multi Admin Approval is now also enforced on Microsoft Graph API calls made with app-only tokens. Until now, only interactive (delegated) admin actions were intercepted. From now on, service principals, PowerShell scripts and third-party tools that change protected resources like apps or scripts are intercepted too. If they don't handle the new approval flow, they fail with an HTTP 403 error. In this blog post I explain what changed, why your automation suddenly gets 403 errors, how the approval flow works for API calls, and how you can exclude your automation apps from the enforcement. An AI agent that works in the playground is maybe twenty percent of the job. The other eighty percent start when real users hit it and you need to answer questions like: why did this run take 40 seconds, which tool call failed, and did the answer quality drop since Tuesday? In this blog post I do a deep dive into observability for agents in Microsoft Foundry: tracing, monitoring dashboards, continuous evaluation on live traffic and alerting. This is the production counterpart to my post about evaluating AI agents in Microsoft Foundry — that one is about testing before you ship, this one is about watching after you ship. In this blog post I explain how I use a Microsoft Foundry Realtime voice agent to talk about information that is currently visible on my screen. I share an Azure page, ask a question with my voice, and receive a spoken answer without copying text into a chat. The useful part is the combination of both inputs. The agent can hear my question and see the screen I selected. I can then ask a follow-up question in the same conversation. The complete example is available in my Microsoft Foundry examples repository.
Read Fabric Data Agents: Chat With Your Data in OneLakeFabric Data Agents: Chat With Your Data in OneLake
Read Agent Skills vs MCP: When to Use Which for AI AgentsAgent Skills vs MCP: When to Use Which for AI Agents
Read Claude Code in GitHub Actions: Automate Standard TasksClaude Code in GitHub Actions: Automate Standard Tasks
Read Multi-Agent Orchestration in Microsoft Foundry: A Deep DiveMulti-Agent Orchestration in Microsoft Foundry: A Deep Dive
Read My Agentic OS: How I Run Notion, Outlook, Codex and OneDriveMy Agentic OS: How I Run Notion, Outlook, Codex and OneDrive
Read Security Copilot Agents in Intune: What You Get in 2026Security Copilot Agents in Intune: What You Get in 2026
Read Microsoft Foundry Model Deployment: A Practical GuideMicrosoft Foundry Model Deployment: A Practical Guide
Read Foundry IQ Deep Dive: Knowledge Bases for AI AgentsFoundry IQ Deep Dive: Knowledge Bases for AI Agents
Read Microsoft Foundry Setup: Project, Model and DeploymentMicrosoft Foundry Setup: Project, Model and Deployment
Read Intune Multi Admin Approval Now Enforced on Graph API CallsIntune Multi Admin Approval Now Enforced on Graph API Calls
Read Microsoft Foundry Observability: Trace AI Agents in ProductionMicrosoft Foundry Observability: Trace AI Agents in Production
Read Microsoft Foundry Realtime Voice Agent: Talk to Your ScreenMicrosoft Foundry Realtime Voice Agent: Talk to Your Screen
Complete archiveAll articles by year
2026
2025
2024
2023
2022
2021
Instruction pages