Microsoft has another free, self-paced AI course, intended for beginners. This one comprises 10 lessons that cover the fundamentals of building AI Agents.

Agents, agents, everywhere - as if LLMs were not enough by themselves! The truth is that agents give superpowers to LLMs, augmenting them by giving them access to tooling. Use cases can be any number of things:

• create personalized itineraries and custom travel handbooks
  
  
• conduct comprehensive research across networks to identify the most suitable suppliers
  
  
• do stock market analysis
  
  
• scrape the web for specific information and present it to the user in a given format
  
  
• make a call and book a restaurant

Microsoft's course gives a few more general use cases:

• Travel agent interprets the context of the email and forwards travel complaints to customer service.
  
  
• Travel agent prioritizes routes with significant price changes based on access to historical pricing data.
  
  
• Travel agent books a journey by determining necessary travel arrangements (car, public transit, flights) from the current location to the destination.

While the use cases vary, the most prolific use case of Agents is in Programming as we detailed in "Potpie - Agentic AI On Your Codebase". They allow you to

chat with your codebase or any public codebase for that matter, to explain it, debug it, or do engineering tasks on it like generating test plans.

Note that this guide is an addition to Microsoft's other great beginner courses on  AI. Those were covered in: Microsoft's Artificial Intelligence for Beginners, Microsoft's Generative AI for Beginners and Microsoft's Generative AI for Beginners Version 2. As such the AI Agents course can be seen as a continuation of the Generative AI Course.

The 10 lessons are:

1. Intro to AI Agents and Agent Use Cases
2. Exploring AI Agentic Frameworks
3. Understanding AI Agentic Design Patterns
4. Tool Use Design Pattern
5. Agentic RAG
6. Building Trustworthy AI Agents
7. Planning Design Pattern
8. Multi-Agent Design Pattern
9. Metacognition Design Pattern
10. AI Agents in Production

As a prerequisite you need a Github oauth token, such as a Github account, since the course uses the Github Models Marketplace to offer free access to Large Language Models (LLMs) that will be used to create our AI Agents. That aside you also need to login to a free Azure OpenAI account which you authenticate with a Microsoft Entra ID.

The code examples are of course in Python and make use of the following Agentic frameworks:

• AutoGen
• Semantic Kernel
• Azure AI Agent Service

Content wise, it's pretty detailed and perfect for beginners. It goes from the bare basics of the Agent's own presence and where they can be used for, to utilizing frameworks and design patterns in building our own and infusing them with RAG. It tackles safety issues, since safety means that the AI agent performs as designed, and then moves on to multi agents and their use cases.

The Metacognition chapter is about the agents' thought process revealing the agent's explicit reasoning about its own reasoning...this goes deep. Finally, the last chapter 10, is about AI Agents in Production, which goes through:

• How to plan the deployment of your AI Agent to production effectively.
  
  
• Common mistakes and issues that you may face when deploying your AI Agent to production.
  
  
• How to manage costs while still maintaining the performance of your AI Agent.

The course keeps good balance between theory and practice. Next to the text are videos provided for the more visually-inclined students. 

All in all, this is yet another stellar offering by Microsoft. Make sure to check it out.

More Information

AI Agents for Beginners - A Course

Related Articles

Potpie - Agentic AI On Your Codebase

Microsoft's Generative AI for Beginners Version 2

Microsoft's Generative AI for Beginners

Microsoft's Artificial Intelligence for Beginners

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook or Linkedin.

Comments

Make a Comment or View Existing Comments Using Disqus

or email your comment to: comments@i-programmer.info