NEW Microsoft AI Agent is INSANE (FREE!) 🔥

Key Takeaways

Microsoft just dropped something insane. You can now build AI agents that actually code for you. It's completely free and open source. I'm going to show you how to set this up in literally two minutes. This changes everything for developers. They took their two biggest AI projects and merged them into one. You get enterprise features and simple setup. No PhD required. No cloud costs if you don't want them. Just pure agent building power. I'm going to walk you through a live demo in VS Code. You'll see how to install it, write the code, and run your first agent. By the end of this video, you'll know exactly how to automate your workflows. This is the future of AI development. Let's go. Hey, if we haven't met already, I'm the digital avatar of Julian Goldie, CEO of SEO agency Goldie Agency. Whilst he's helping clients get more leads and customers, I'm here to help you get the latest AI updates. So, Microsoft just released something called the Microsoft Agent Framework. And honestly, this is wild. This is a public preview that just went live. It's open source. It's on GitHub right now. You can grab it today. And what it does is lets you build AI agents that can actually do work for you. We're talking coding agents, automation agents, agents that can handle your workflows. And the crazy part is how easy this is to set up. Now, before I dive into the demo, let me tell you what makes this different. Microsoft took two of their biggest AI projects, Autogen and Semantic Kernel, and merged them together. AutoJem was all about agent orchestration, how agents talk to each other. Semantic kernel was the enterprise stuff, the production ready features. Now you get both in one framework. And here's why that matters. You're not just getting a toy to play with. You're getting something that scales, something that works in production, something that integrates with Azure if you need it, but you can also run it locally for free. The framework supports Python and it's MIT licensed. That means you can use it for anything. Personal projects, commercial projects, whatever you want. Now, let me tell you what you can actually build with this thing. You can build coding agents that write code for you, review your code, fix bugs. You can build automation agents that handle ticket triage, customer support, content generation, multi- aent workflows where different agents handle different parts of a bigger task. And the best part is the framework handles all the hard stuff for you. the orchestration, the state management, the tool calling, you just focus on what you want the agent to do. All right, so I'm going to show you exactly how to get started. I'm going to do a live demo in VS Code. We're going to install this thing, scaffold a basic agent, run it, see it actually work. But before we jump into the code, let me tell you about my AI profit boardroom. This is where I teach people how to scale their business with AI automation. We're talking saving hundreds of hours, getting more customers, automating the boring stuff. If you want to actually make money with AI instead of just playing with it, you need to check this out. The link is in the description. All right, demo time. The first thing you need is Python 3.10 or higher. I'm in VS Code right now. I've got the Python extension installed. The repo is github.com/microsoft/ aent-framework. You can grab it right now. Go start it while you're there. Now, I'm going to create a new virtual environment. I'm opening my terminal in VS Code. I'm typing python-mv. That creates the environment. Then I activate it. On Mac or Linux, you do source venv/bin/activate. On Windows, you do venv back slashscripts back slactivate. Now, here's where it gets interesting. To install the framework, you run pip install agent-framework. Heads up, this is a public preview, so sometimes you might need to access GitHub packages. The docs will tell you if you need a personal access token. Just follow the quick start instructions. All right. Now I've got the framework installed. Time to write some code. I'm creating a new Python file. Let's call it demo. py. So here's what I'm writing from agent framework import agent, tool. Now I'm defining a simple tool, just a function def tool with input as a parameter and it returns result plus the input. Super basic. This tool just echoes back whatever you give it. Now I create the agent. agent equals agent with name equals demo- agent comma tools equals a list with tool the name is echo and the function is my underscore tool then I tell the agent to run response equals agent run with a prompt like make a to-do list for launching a blog and then print the response that's it that's the entire agent like 15 lines of code now let me run this thing the agent processes the request it plans out the steps it calls the tool if it needs to and it gives me back a response and here's the thing this demo is super basic but You can scale this up. You can add more tools. You can have multiple agents working together. You can integrate it with APIs, databases, external services. Now, let's talk about scaling this thing because running agents locally is cool. But what if you want to deploy this in production? That's where Azure AI foundry and the agent service come in. Azure AI Foundry is Microsoft's platform for building and deploying AI apps. It's got agent hosting built in. You can deploy your agents there, scale them up or down based on demand, monitor performance, track costs, and because the agent framework integrates with Azure, it's basically plugandplay. You build locally, you test locally, then when you're ready, you push to Azure, and it just works. Now, let's talk about security. When you're running agents, especially agents that call external tools or APIs, you need to think about permissions, about tokens, about data boundaries. The agent framework is in public preview. That means APIs might change. You need to stay close to the docs and the GitHub repo. Test everything in a safe environment before you go to production. When your agents call external services, make sure you understand what data they're accessing. Use API keys wisely. Don't hardcode secrets. Use environment variables. And remember, agents are powerful. They can do a lot. That means they can also make mistakes. Always test in isolated environments first. All right. So, where do you go from here? First, start the repo on GitHub. It's github.com/microsoft/agent-framework. This helps Microsoft know people care. Second, check out the samples in the repo. There are Python examples,Net examples, different use cases. These are gold. Third, read the docs. Microsoft has a full quick start guide on their learn platform. It walks through installation, setup, your first agent, everything you need to get started. Now, let me tell you why this matters for you. If you're building AI products, you need agents. Simple prompts aren't enough anymore. You need systems that can plan, that can use tools, that can break down complex tasks. This framework gives you that for free with Microsoft backing it. If you're an SEO agency owner like me, think about what you could automate. Content generation, link prospecting, outreach, reporting, all of that could run through agents. You build it once, it runs forever. That's leverage. If you're a developer, this is a skill you need. Aentic AI is the next wave. Companies are going to need people who can build this stuff. Get in early. Learn it now. All right, so let's wrap this up. Microsoft Agent Framework, public preview, open-source, Python, andnet. You can build agents locally or deploy to Azure. It's easy to get started. The docs are solid. This is one of those releases that changes the game. Not because it's magic, but because it makes hard things easy. You don't need a PhD to build agents anymore. You just need this framework and some basic coding skills. So go grab it, build something, break something, learn. Now before you go, let me tell you about the AI money lab. This is my free community for anyone who wants to make more money with AI. Inside you get 50 plus free AI tools, 200 plus chat GPT SEO prompts, over a,000 free and 8N workflows, a full blueprint to generate thousands of leads with AI completely free. You'll learn how to make money with AI agents, how to automate your business, how to scale without hiring more people. We've got case studies, we've got templates, we've got a community of people doing this stuff every day. The link is in the description. Join for free. No catch, just pure value. And if you want to go deeper, if you want to actually scale your business with AI, check out the AI Profit Boardroom. This is my paid community where I teach the advanced stuff. How to build AI systems that actually make money, how to save hundreds of hours with automation, how to get more customers using AI. The link is also in the description. Julian Goldie reads every comment, so drop a comment below. Tell me what you think about this framework. Tell me what you're going to build. Ask questions. I'll answer them. And if you got value from this video, smash that like button. Subscribe if you haven't already. Hit the bell so you don't miss the next update. All right, that's it for today. Go build something wild.