AI-901: Azure AI Fundamentals

Explore the Azure global backbone, with 60 regions, fiber and subsea cables, edge sites, and peering to deliver low-latency services; learn region pairs and availability zones for redundancy.

Explore the four Azure subscription types—free, student, pay-as-you-go, and enterprise agreement—and learn when each applies for learning, development, and business needs.

Create your Azure subscription by selecting the free option or pay-as-you-go, sign in with your Microsoft account, and access 12 months free, 55 services always free, and a $200 credit.

Discover how to set up and explore your Azure environment. Inspect the resource hierarchy, including management groups, subscriptions, and resource groups, and learn to create and move them.

Learn how to configure a budget for your Azure subscription in cost management, including scope selection, monthly thresholds, and alerting based on actual or forecasted spend.

Compare generative AI with traditional AI, highlighting content creation over data analysis, and explore use cases in text, images, and code generation based on data sources and prompts.

Explore how large-language models predict the next token from context, rather than accessing knowledge. See how prompts are processed by neural networks and transformers and how training data shapes output.

Compare generative AI with agentic AI, showing how agentic AI can autonomously or semi-autonomously execute tasks such as autonomous coding, end-to-end ticket resolution, and multi-agent workflows within defined boundaries.

Explore five AI model types—large language models, code models, diffusion models, multimodal models, and domain-specific models—and their use cases with examples like GPT-4, GPT-5, Gemini, codecs, and Security Co-Pilot.

Microsoft Foundry is a Gen AI first platform for building and deploying AI apps and agents, while Azure Machine Learning centers on data prep, model training, and MLOps.

Foundry pricing is complex and varies by service under a pay-as-you-go, token-based model. Costs span models, agent services, knowledge and tools, observability, trust, and local deployments, requiring monitoring at scale.

Deploy a Foundry environment in Azure by provisioning a resource and project, and configuring networking, identity, and encryption. Then access the Foundry portal to explore models, tools, templates, and guardrails.

Deploy an OpenAI model in the Foundry portal by selecting GPT versions, configuring deployment options and guardrails, and understanding token usage.

Explore the Foundry agent service as the assembly line for building agents, integrating inputs, tools, multi-agent workflows, and Azure Monitor with secure, role-based access control.

Create an it helpdesk triage agent in the Foundry portal for Azure AI Fundamentals. Configure instructions, a web search tool, and guardrails, then test and publish.

Deploy your AI agent in Microsoft Foundry using Python and a GPT 4.1 mini model, creating a fitness coach that offers workouts and tips while handling environment and conversations.

Configure an Azure AI speech resource to convert text to natural speech and transcribe audio back to text, using SSML, the Speech SDK, and Foundry tools in a Python notebook.

Architect and use Azure Document Intelligence to read PDFs, extract invoice and receipt data, detect handwritten text, and deploy the resource with ARM templates and RBAC for end-to-end document processing.