AI901: Microsoft Azure AI Fundamentals Full Course Exam 2026

Introduce artificial intelligence and intelligent automation, using a toll plaza analogy, and highlight AI, ML, and DL, with examples like facial recognition and recommendations.

Explore how machine learning replaces manual rules with data-driven models that learn patterns and make predictions. Learn the 3-step pipeline—collect data, train the model, evaluate results and improve—through real-world examples.

Explore the two main machine learning models—supervised and unsupervised—learning from labeled or unlabeled data, including classification, regression, clustering, association, and dimensionality reduction, with examples like spam detection and weather prediction.

Compare predictive AI and generative AI to show how historical data and models forecast outcomes. Explain how prompts generate new content and outline key use cases and outputs.

Explore AI speech technology, including speech recognition and synthesis, enabling voice commands, real-time transcription and captions, voice assistants, and translation across languages in voice-enabled apps.

Explore the six-step speech recognition pipeline, from audio capture to post-processing, and see how acoustic and language models decode speech into accurate text.

Explore the speech synthesis workflow as an assembly-line process, moving from normalization to linguistic analysis, rosodi generation, and audio synthesis to convert text into natural speech.

Explore natural language processing (NLP) and its core tasks, including language detection, classification, sentiment analysis, entity extraction, key term extraction, PII detection, and text summarization, with real-world use cases.

Explore the NLP workflow, starting with tokenization and normalization, then stop word removal, stemming or lemmatization, pause tagging, and ngrams to build a robust NLP pipeline.

Explore computer vision, a branch of AI that teaches computers to see, understand, and decide from images and videos, with examples in self driving cars and smart retail.

Explore four core computer vision tasks, image classification, object detection, semantic segmentation, and contextual image analysis, and see how they power modern AI applications.

Explore convolutional neural networks for image processing that learn visual features like edges, shapes, colors, and textures through labeled training data and backpropagation, enabling face recognition and medical imaging applications.

Explore how transformers use self-attention to understand context. Learn how visual transformers divide images into patches with embeddings and how multimodal models fuse image and text for tasks like captioning.

Discover how ai-powered information extraction turns unstructured data into structured insights from documents, images, audio, and video. Learn how ocr and speech-to-text enable practical use in finance, legal, and healthcare.

Explore how optical character recognition (OCR) converts image text into editable, searchable text through a five-stage workflow: image capture, pre-processing, region detection, character recognition, and output generation, using artificial intelligence.

Explore how OCR differs from field extraction by identifying critical data points and mapping them to structured invoice fields. Learn how data ingestion, standardization, and storage enable automated document processing.

Explore generative ai, the technology that creates new content from prompts, including text, images, and code. Learn its use cases from chatbots to personalized learning, and compare LLMs and SLMs.

Explore how large language models work, from input to tokens, word embeddings, and next-word prediction. See a step-by-step example with 'i am eating' to illustrate outputs.

Explore ai hallucinations in generative ai, learn why language models sound confident yet may be false, and apply retrieval augmented generation, prompts, and verification to reduce errors.

Explore responsible AI by examining its six pillars, such as fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. Learn how ethical, trustworthy AI benefits society.

Explore scenarios of responsible AI that illustrate fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability, including governance oversight and human in the loop considerations.

Explore how Azure regions cluster data centers to deliver low latency by placing resources near users, and how global expansion enables scalable, worldwide deployment.

Explore Azure resources and resource groups, learn how resources are created and organized into a logical resource group to prevent cloud chaos, with rules on movement, containment, and governance.

Azure subscriptions are logical billing and access boundaries that act as management and scale units, linking accounts to resource groups and enabling dev, test, and prod isolation.

Explore how Microsoft Foundry acts as a factory for AI applications, using models, agents, tools, and knowledge in an agent-first architecture to build, test, and deploy enterprise AI at scale.

Create a Foundry AI project, deploy models, and build agents while exploring the Foundry portal, playground, model catalog, and guardrails to manage API access and project endpoints.

Explore the Foundry model catalog to filter and compare Azure and partner models by source, capability, region, and task, then deploy, monitor, and manage costs via API endpoints.

Deploy an AI agent in Microsoft Foundry by configuring a travel-assistant with a GPT-4.1 model, instructions, and tools to plan trips and answer travel questions.

Explore a journal-purpose AI model for text analysis, showing key phrase extraction, entity linking, sentiment analysis, opinion mining, and summarization with no configurations or training.

Demonstrates speech recognition and speech synthesis using Azure Speech in a playground, converting spoken words to text and text to speech with adjustable voices and emotional expression.

Deploy a speech capable agent to conduct real time spoken conversations, enabling speech to speech interactions such as voice assistants and speech translation, then explore Microsoft Foundry.

Deploy a text-to-image generation model using PLUX 1.1 Pro on Microsoft Foundry, prompt a description like 'tiger walking', and view sample code to implement programmatically.

Learn how to extract information from invoices and other documents with Azure content understanding using a base model and pre-built analyzers to convert unstructured data into structured json for storage.