AzureFeeds

This week in Model Mondays, we dive into the cutting-edge research happening at Azure AI Foundry Labs. From the MCP Server that makes it easy to experiment with new models and tools, to Magentic-UI that brings human-centered agent workflows to life, there’s a lot to unpack!

Spotlight On: Research & Innovation in Azure AI Foundry Labs

1. What is this topic and why is it important?

Research and innovation are the engines that drive progress in AI. This episode spotlights Azure AI Foundry Labs—a hub where Microsoft’s latest breakthroughs in models, agent frameworks, and developer tools are shared with the world. The MCP Server makes it easy for an

yone to experiment with new models and tools, while Magentic-UI brings human-centered agent workflows to life. These projects help bridge the gap between cutting-edge research and real-world applications, making it easier for developers and students to try new ideas and build smarter AI solutions.

2. What is one key takeaway from the episode?

The key takeaway is that Azure AI Foundry Labs is not just a showcase—it’s a hands-on playground. You can access open-source projects, sample code, and demos, and even deploy models directly in your Azure workspace. The MCP Server and Magentic-UI are designed to make experimentation fast and accessible, so you can go from idea to working prototype in minutes, not months.

3.  How can I get started?

• Visit Azure AI Foundry Labs to explore Foundry Labs projects and resources.


• Try out the MCP Server by following the quickstart guides and sample repos.


• Experiment with Magentic-UI to see how multi-agent systems can automate web tasks with you in the loop.


• Check out the GitHub repos for sample code, documentation, and community contributions.

Highlights: What’s new in Azure AI Foundry?

• Zero Trust Agents: Microsoft is bringing identity and access management to agentic workflows. Each AI agent can now have its own badge, credentials, and security clearance using Entra ID. This means more secure, auditable, and professional AI operations—no more rogue bots!


• Trellis Model: Foundry Labs launched Trellis, a model that creates high-quality 3D assets from text or image inputs. It uses a unified latent space and delivers detailed, editable 3D models for gaming, virtual worlds, and industrial design.


• ISO 421 Certification: Azure AI Foundry and Microsoft Security Copilot are now ISO 421 certified, the gold standard for AI management systems. This ensures your AI deployments are secure, compliant, and globally recognized.


• Fine-Tuning Updates: Azure AI Foundry now supports preference alignment, global data training, and more control over SFT and DPO blends. Models can be tuned for accuracy and user preferences, with multilingual and multicultural support.


• Kubernetes AI Assessment Tool: Microsoft launched an AI-powered tool to discover, diagnose, and document Kubernetes clusters. It provides human-readable summaries, making DevOps easier and reducing outages.


• Foundry Labs: The hub for all new research, models, and tools. You can find open-source projects, sample code, and demos, and deploy models directly in your Azure workspace. The MCP Server and Magentic-UI are two key projects making experimentation fast and accessible.

What is Azure AI Foundry? (Simple Explanation for Students)

Azure AI Foundry is like a playground and toolkit for anyone interested in building AI solutions end-to-end. Imagine a place where you can find the newest AI models, tools, and research from Microsoft, and actually try them out yourself. Whether you’re a student, developer, or researcher, Foundry helps you learn, experiment, and build cool AI projects—without waiting for products to be released. It’s all about making innovation hands-on and accessible!

Foundry Labs: is a dedicated space within Azure AI Foundry, to showcase research models and tools. The MCP Server and Magentic-UI are two of the key projects here, that we focused on in this episode.

Demos & Real-world Applications

One of the best parts of this episode was seeing real demos and practical uses of Azure AI Foundry Labs projects:

• MCP Server Demo: The MCP Server lets you quickly set up and experiment with new models and tools. You can run it locally or in VS Code, and use it to build your own agents or workflows. The demo showed how easy it is to start from a template and customize your own AI project.


• Magentic-UI Demo: Magentic-UI is a multi-agent system that can automate web tasks, generate and execute code, and interact with websites—all while keeping you in control. The demo walked through how agents plan, act, and ask for your approval before doing anything sensitive. You can use it for things like booking appointments, finding travel deals, or even building a custom pizza online!


• OmniParser Demo: OmniParser is a vision module that turns UI screenshots into a JSON map of clickable elements. The demo showed how it can help agents “see” and understand what’s on the screen, making them smarter and more useful for real-world tasks.


• Real-World Example: The episode also featured a customer story from the Capacity team, who use Azure AI Foundry to build an answer engine that helps organizations find knowledge across all their data sources. Their platform uses Microsoft AI to break down data silos, empower people, and deliver secure, enterprise-grade solutions.

Why does this matter?

These demos and stories show that Azure AI Foundry isn’t just for research—it’s being used to solve real problems, build useful apps, and help people in their daily work. Whether you want to experiment, learn, or create something new, Foundry gives you the tools and resources to make it happen.

Writing Weekly Blogs: Sharda’s Tips

As a Model Mondays blogger, I author a new blog post every week, on a new AI model, tool, or concept. Writing a technical blog like this is all about breaking down complex ideas and making them accessible for everyone, especially if you’re a student or developer. Here’s how I approach it:

• Start with the technical core: I focused on the MCP Server, Magentic-UI, and OmniParser, explaining how each works and why it matters for real-world AI development. I used details from the transcript to highlight demos and practical applications.


• Connect research to practice: I made sure to show how Azure AI Foundry bridges the gap between research and hands-on experimentation, so you can actually use the latest models and tools in your own projects.


• Keep it simple but precise: Even though the topic is technical, I used clear language and step-by-step explanations, just like a teacher would in class.
  Share your workflow: I used GitHub Copilot and the following prompt to help structure and draft this blog. You can use it too!

Prompt I Used:

Generate a technical blog post for Model Mondays S2E6 based on the transcript and episode details. Focus on MCP Server, Magentic-UI, OmniParser, and real-world demos. Explain Azure AI Foundry for students, add a section on practical applications, and share tips for writing technical blogs. Make it clear, engaging, and useful for developers and students.

How did it work?

Using this prompt with Copilot made it much easier to organize my thoughts, pull in technical details, and keep the blog focused on what matters for developers. Try using this prompt and adjust it for your topic!

Coming Up Next Week

Next week, we’ll explore Model Auto-Selection with the Model Router—a deployable chat model that intelligently routes requests to the best available model in real time. If you’ve ever wondered how to pick the right model for your app, this episode will show you how automation can make model selection smarter and easier. Register Here!

About Model Mondays

Model Mondays is a weekly series designed to help you build your Azure AI Foundry Model IQ with three elements:

• 5-Minute Highlights – Quick news and updates about Azure AI models and tools on Monda


• 15-Minute Spotlight – Deep dive into a key model, protocol, or feature on Monday


• 30-Minute AMA on Friday – Live Q&A with subject matter experts from Monday livestream

Want to get started?

1. Register For Livestreams – every Monday at 1:30pm ET


2. Watch Past Replays to revisit other spotlight topics


3. Register For AMA – to join the next AMA on the schedule


4. Recap Past AMAs – check the AMA schedule for episode specific links

Join The Community

Join the Discord – for real-time chats, events & learning

Explore the Forum  – for AMA recaps, Q&A, and help!