I asked ChatGPT about the course, response below.
The reason i want to know, because if this 'Microsoft Foundary' isn't python, its not going to help me.
I don't know what Microsoft Foundary is..
===My Question to ChatGPT
my question, does this include python in building of AI Agents? is this a marketable skill to learn
Microsoft Foundary?
"AI Agents for Beginners 18 lessons * Build AI Agents with Microsoft Foundary Now with Deploying Scalable * Local AI Agents Microsoft dropped this FREE 18 lesson course to learn about AI Agents (from Scratch) Code + Videos = Complete Learning!"
**Response below from ChatGPT
Yes. The course you found is
very relevant if you want hands-on Python experience building AI agents, and
Microsoft Foundry is a marketable platform skill—especially when paired with Python and agent testing/evaluation. 
Microsoft’s current repository confirms that the course uses
Python 3.12+, Jupyter notebooks, Microsoft Agent Framework (MAF), and Microsoft Foundry Agent Service V2. Microsoft describes the lessons as containing
Python code samples using Microsoft Agent Framework with Microsoft Foundry.
What you'll actually be learning
It isn't just clicking around in a Microsoft cloud interface. You'll be working with a stack roughly like this:
Python
↓
Microsoft Agent Framework
↓
LLMs / tools / APIs / RAG / memory
↓
Microsoft Foundry
↓
Deploy + evaluate + monitor agents
The setup specifically has you create Python virtual environments and install Python dependencies, and the exercises are supplied as Python Jupyter notebooks.
More importantly, the curriculum covers actual agent concepts such as
tool use, planning, multi-agent workflows, RAG, memory, evaluation, human approval, observability, deployment, and local agents.
Is Microsoft Foundry marketable?
Yes—but I would not learn "Foundry" in isolation.
The more marketable combination is:
Python + AI agents + APIs + RAG + evaluation/testing + Microsoft Foundry/Azure
That combination lets you describe yourself as someone who can actually
build and test agentic applications,
[YES - BlackRob] rather than someone who merely knows how to navigate Foundry.
This is particularly useful because Microsoft positions Foundry Agent Service as the managed runtime while Microsoft Agent Framework provides the Python/C# SDK for agents, tools and workflows.
And there's a very useful advanced lesson for you: Microsoft's July 2026 update added
Deploying Scalable Agents with Microsoft Foundry, including tools, RAG, memory, model routing, caching, human approval, an
evaluation gate, and OpenTelemetry tracing.
That
evaluation/testing component is especially valuable for someone approaching AI from a software-quality background.
How I'd fit this into your learning path
You don't need to abandon LangGraph or OpenAI agent development. Think of them as complementary skills:
| Skill | What you're learning |
|---|
| Python | Core programming language |
| OpenAI API | Directly building LLM applications |
| LangGraph | Agent orchestration/workflows |
| Microsoft Agent Framework | Microsoft's agent framework |
| Microsoft Foundry | Enterprise platform for building/deploying/managing AI |
| RAG | Connecting agents to enterprise knowledge |
| Pytest / evaluation | Testing whether agents work correctly |
| Azure | Cloud infrastructure/ecosystem |
So I would
absolutely take this free course.
The official Microsoft course is here:
Microsoft AI Agents for Beginners GitHub course
One thing to note: you may see older descriptions saying
12 lessons. Microsoft's repository documentation now describes the curriculum as
18 lessons numbered 00–18, and the July 2026 update added scalable Foundry deployment and local-agent material.
After you've completed several of the Python labs, I'd also recommend putting a small
Microsoft Foundry AI Agent project on GitHub rather than merely listing "Microsoft Foundry" under Skills. That gives you something concrete to discuss in AI QA, SDET, AI Test Engineer, or Agentic AI interviews.
docs.cdp.coinbase.com
