Working with large language models using Azure

Overview

Learn how to build, customize, and deploy generative AI applications using Large Language Models (LLMs) and Microsoft Azure. This hands-on course introduces the practical techniques developers use to improve AI application performance, reliability, and business relevance. You’ll begin by exploring how LLMs work, including their architecture, capabilities, and limitations. From there, you’ll apply prompt engineering strategies to improve model outputs and build more effective AI interactions. The course then introduces Retrieval-Augmented Generation (RAG) pipelines, teaching you how to connect LLMs with external data sources to deliver grounded, accurate responses. You’ll also learn how to customize models using fine-tuning techniques and evaluate when to use fine-tuning, RAG, or hybrid approaches for different business scenarios. In the final modules, you’ll build and deploy generative AI applications using Azure AI Foundry and Azure OpenAI services while learning deployment, monitoring, and cost management strategies. By the end of this course, you’ll have practical experience building AI-powered applications using modern Azure AI tools and workflows.

Syllabus

Understanding Large Language Models (LLMs)

This foundational module introduces the core concepts behind Large Language Models (LLMs). You will start by exploring the fundamental architecture that powers models like GPT (Generative Pre-trained Transformer) and learn how they process information and generate human-like text. The second half of the module is dedicated to prompt engineering, where you will learn and apply essential techniques—from basic commands to advanced strategies like few-shot learning and chain-of-thought—to effectively communicate with and control AI models to achieve desired outcomes. Important Notice on the Azure Interface: The screencast videos and screenshots were last updated in late 2025. Please be aware that Microsoft may have updated the Azure interface since then. If the steps shown in the course materials look different from your current Azure environment, please follow the most up-to-date interface, as the underlying concepts and learning objectives remain the same.

Implementing RAG pipelines

This module focuses on one of the most powerful techniques for enhancing LLMs: Retrieval-Augmented Generation (RAG). You will learn how to ground models in external, private, or real-time data sources to provide more accurate and contextually relevant responses. You will start by building a basic RAG pipeline using Azure services and then progress to constructing and optimizing advanced systems with techniques like semantic ranking and sophisticated data chunking strategies. Important Notice on the Azure Interface: The screencast videos and screenshots were last updated in late 2025. Please be aware that Microsoft may have updated the Azure interface since then. If the steps shown in the course materials look different from your current Azure environment, please follow the most up-to-date interface, as the underlying concepts and learning objectives remain the same.

Fine-tuning and customizing LLMs

This module explores fine-tuning as a powerful method for customizing an LLM's core behavior, style, or knowledge for specialized tasks. You will learn the entire fine-tuning workflow, from preparing a high-quality dataset to launching the training job and evaluating the customized model's performance in Azure. Critically, you will learn to strategically decide when to use fine-tuning versus RAG—or a hybrid of both—to create highly effective, domain-specific AI solutions. Important Notice on the Azure Interface: The screencast videos and screenshots were last updated in late 2025. Please be aware that Microsoft may have updated the Azure interface since then. If the steps shown in the course materials look different from your current Azure environment, please follow the most up-to-date interface, as the underlying concepts and learning objectives remain the same.

Developing generative applications with Azure

This module transitions from theory to practice by guiding you through the end-to-end process of building and deploying a complete generative AI application. You will learn to design an application's architecture and user flow before using Azure AI Foundry and Prompt flow tools to build it. The module then covers the critical MLOps lifecycle, teaching you how to deploy your application as a secure endpoint, manage it in a production environment, and implement monitoring with Azure Monitor for performance and cost. Important Notice on the Azure Interface: The screencast videos and screenshots were last updated in late 2025.Please be aware that Microsoft may have updated the Azure interface since then. If the steps shown in the course materials look different from your current Azure environment, please follow the most up-to-date interface, as the underlying concepts and learning objectives remain the same.