Azure AI Foundry Mastery + LLMOps and More

Name: Azure AI Foundry Mastery + LLMOps and More
Rating: 3.8 (35 reviews)

Azure AI Foundry, Generative AI, LLMOps

Created bySteven Rodriguez

Last updated 6/2025

English

What you'll learn

Understand the basics of Azure AI Foundry and its components
Gain practical experience in Generative AI through use-cases and leveraging Azure AI Studio
Explore how to build, train, and deploy AI models using Azure AI Studio
Learn best practices for managing resources in Azure AI Studio and Integration

Course content

4 sections • 42 lectures • 6h 20m total length

Introduction to Course4:29
Introduction to the course and course outline to ensure each section is updated as of November 22, 2024 this reflects name changes associated with Azure AI Foundry.
What is Azure AI Foundry?10:59
[HOL] Creating Azure AI Studio in Azure (Console)5:53
Creating Azure AI Studio from the console view this will show you what resources and configuration items are included in the Wizard. Inside the source code in the resources of this course will have this in Azure CLI if you'd opt to programmatically apply resources.
Creating Azure AI Studio with Private Access3:07
Refresher: Components behind Azure AI Studio (Conceptually for beginners)4:14
Refresher content for those unfamiliar with Microsoft Azure and related components that make up the Azure AI Studio construct.
RBAC Roles in Azure AI Studio9:18
Working with Azure AI Studio uses RBAC for controlling access, this lecture covers the underlying permissions required to create, maintain and also use Azure AI Studio and related components.
Data Dictionary and Course Terms (Explained)14:34
A data dictionary of terms that are used throughout this course are discussed in categories that are attributed to the scope of the course. These terms are used throughout the course and if you are new to using AI and Azure Cloud this is intended to assist you.
Best Practices for Azure AI Foundry (Security Baseline) Part 112:14
Azure AI Foundry has a security baseline defined by Microsoft these are specific controls related to components of the service and are categorized to help customers identify specific parameters and features to secure the service when utilized in production and in general best practices of the service.
Best Practices for Azure AI Foundry (Security Baseline) Part 210:58
Azure AI Foundry has a security baseline defined by Microsoft these are specific controls related to components of the service and are categorized to help customers identify specific parameters and features to secure the service when utilized in production and in general best practices of the service.
Bing API Changes and Updates3:35
Bing Search API that has been used in some of the demos such as Azure OpenAI + Bing Search API are being deprecated to use Grounding with Bing Search this introduces some new price changes therefore this video outlines moving forward the labs with this API will now support and use Exa.AI.
Azure Foundry Updates [June 2025]6:29
Azure Foundry has added/upgraded various services and offerings in preview this video is to give a high-level overview of these changes to ensure this course keeps up with the innovation.

Source Code and Resources0:11
Source code associated with this course will be exclusively hosted on GitHub via the following link.

Feel free to fork/clone this repo to use for each lab
[HOL] Getting Started Lab (Python)9:08
This lab introduces the use of SDK in Python for sending queries into Azure OpenAI with a simple construct to get started, this shows the output in raw format in JSON and also with it cleaned up to the end user similar to how the experience is in the Chat Playground.
RAG (Retrieval Augmented Generation) in Microsoft Azure14:54
RAG Continued (Patterns/Types)13:24
Explore the fundamentals of Retrieval-Augmented Generation (RAG), a powerful AI approach that combines retrieval systems with generative models to enhance accuracy and relevance. This video breaks down key concepts, workflows, and practical use cases to help you harness RAG for advanced AI-driven applications.
[HOL] AI Search with Generative AI11:11
Use of Azure AI Search Bring your Own Data.
This does consumes resources in azure such as Azure AI Search, Azure OpenAI (Embedding Model), Azure OpenAI API Calls.
Deploying a Chat Application6:26
This video demonstrates the use of the Chat Application via Microsoft's deploy to web app that using the App Service (PaaS) solution this also has the capability to use Azure Cosmos DB for logging chat interactions for monitoring and uses the Entra ID (IdP) solution for authorization/authentication of users.
[LAB-5] Use of Jina AI for Summarization on Azure OpenAI9:16
Jina.AI is a AI Search/Embedding service that streamlines the use of being able to provide a link to a site and summarize the findings or extract the findings in a markdown format that is beneficial to a Large Language Model's existing ability to understand the subject and provide context to the end user.
[LAB-6] Infrastructure as Code Deploying Azure OpenAI9:30
Infrastructure as code or commonly (IaC) is the automation of creation and maintaining infrastructure via code in cloud environments this is commonly used by multiple programming languages in this video I demonstrate the use of Terraform for deploying an Azure OpenAI model.
[LAB-9] Function Calling in Azure OpenAI with Bing API15:00
Function calling extends the capabilities of using Generative AI to other functions such as tools that can assist in a better user experience. For components that knowledge might be limited a example in this video demonstrates the use of Bing Search API to continue a conversation on context outside of the LLM's existing knowledge base.
[LAB-10] Chainlit Application with Azure OpenAI + Exa.AI15:24
Chainlit is a frontend python library that allows you to have conversations with a LLM similar to how chat applications work asynchronously, this demo shows how we add the components of search to it by using Exa.AI.
Assistants API7:35
The Assistants API in Azure OpenAI enables users to design and deploy intelligent, task-specific virtual assistants that integrate seamlessly into their workflows. It empowers end users to automate processes, retrieve insights, and interact with AI in a conversational, intuitive way tailored to their unique requirements with use of tools such as code interpreter and function calling.
[HOL] Assistants API Playground with Example5:14
Discover the Assistants API Playground in Azure AI Studio, where you can build, customize, and deploy intelligent assistants tailored to your needs
Azure Foundry AI Agent Service Introduction11:09
Azure Foundry released the AI Agent Service which is a managed service to allow creation of AI Agents with either programmatically via code or use of a Wizard in the foundry. This abstracts to the end user to allow out of the box building such as first party native tooling. This is still in preview and the demo that assists with understanding this code is Lab 15 - Semantic Kernel Part 1/Part 2.

Evaluation Concepts and Components10:48
Evaluations in Azure AI Studio7:17
Adversarial Scenario Simulation for Azure OpenAI11:05
[LAB] Use of DALL-E Model for Image Generation via Code12:37
Responsible AI10:33
Responsible AI ensures the ethical development, deployment, and use of artificial intelligence by prioritizing fairness, transparency, and accountability. This lecture explores the principles and practices for building AI systems that align with societal values and mitigate potential risks.

Prompt flow in Azure AI Studio7:52
[LAB] Prompt Flow UI/UX Experience + Set up of Code12:43
[LAB] Prompt Flow via VS Studio Running Flow8:25
[LAB] 08-Research Flow in Prompt Flow LLMOps18:38
[LAB] 11-Promptflow-Exa (Architecture Part 1)3:40
This video discusses the use of prompt-flow a tool developed and open-sourced by Microsoft along with use of third-party AI Search service Exa.ai to demonstrate how you can use prompt-flow to structure a LLM Input and get responses more accurately. The second portion of this video will show the execution of the code and outputs along with additional information.
[LAB] 11-Promptflow-Exa (Part 2)10:41
[LAB] 12-LlamaIndex with Azure AI Search (Architecture) Part 16:18
This video introduces the constructs of the architecture that is required to get the code in part 2 running efficiently. The use of the following resources are used in this video Chat Completions Model (Gpt-4o-mini), Embed Model (Text-ada-002), Azure AI Search (Free SKU).
[LAB]12-LlamaIndex with Azure AI Search Part 211:02
This video demonstrates using a jupyter notebook located in the 12-llamaindex folder in the repository how to use LlamaIndex with Azure AI Search to retrieve data from our custom data uploaded along with how to query it using query engine.
[LAB]13-Evaluations in Azure Foundry9:41
[LAB]14-Azure AI Inference Part 1 [DeepSeek R1 Setup]2:49
This video shows the setup of using DeepSeek R1 via Azure Foundry. The setup covers the deployment method of a 3P model using Base model then gathering the required parameters that go into the .envexample shown in the repository that sets up for Part 2.
[LAB]14-Azure AI Inference Part 2 [DeepSeek R1]8:55
This lab shows how to run Serverless 3P models such as DeepSeek-R1 and use of Exa.AI with LangChain.
[LAB]15-Azure Foundry Agents Part 1 [Semantic Kernel]6:33
[LAB]15-Azure Foundry Agents Part 2 [Semantic Kernel]6:12

Requirements

Python Familiarity (Fundamentals)
Cloud Computing Concepts
Basic knowledge of data science and machine learning principles

Description

Many organizations are harnessing the power of Generative AI typically from a Software as a service offering such as ChatGPT Enterprise, Microsoft 365 Copilot and Duet AI. What if your use case isn't met by a out-of-box vendor? This is where the use of platforms such as Azure AI Foundry that allow the fine-tuning, evaluation, prompt flow and more components to assist your developers to produce production grade generative AI applications.

If you are a Developer, AI Engineer, Data Scientist or interested in learning what tools and techniques you can use to develop new applications on a platform such as Microsoft Azure this course is for you. Most organizations that are considering using Generative AI want more control over the use of which frontier models, this course intends to expand the conversation to cover what is possible to build on the platform along with the vast array of use of open-source models such as LLaMa-3,Mixtral and more. This course is broken up into sections with a mixture of theory, lectures, and practical hands-on the repository added to this course is publicly accessible for you to replicate the labs with a README attached to each folder to help you step by step.

This course is intended to get you up to speed on the latest use of Generative AI but also how to use Microsoft Azure as your building Generative AI along with introducing the capabilities behind this service that is now Generally Available with hands-on labs demonstrating the capabilities.

Who this course is for:

AI Developers
Software Developers
Data Scientists
AI Enthusiast
Generative AI Engineers

Azure AI Foundry Mastery + LLMOps and More

What you'll learn

Explore related topics

Course content

Introduction to Course11 lectures • 1hr 26min

Section 2: Building with AI Models13 lectures • 2hr 8min

Section 3: Deploying and Managing AI Models5 lectures • 52min

Section 4: Advanced Features and Custom Solutions13 lectures • 1hr 53min

Requirements

Description

Who this course is for: