See inside a real Azure AI lab

Info

You will already have a project when you sign in

Microsoft has changed the first-run experience. Signing in now creates a project for you automatically, so you land on the Home page of a project you never asked for, with an API key and endpoints already filled in.

Two things follow from that, and which one applies depends on the lab.

Some labs in this course provision a project for you with a specific name. Those labs say so at the top and tell you to work inside it.

This lab is not one of them. Here you create your own project on purpose, and work in that. Follow the steps below exactly, even though a project already exists when you arrive. Creating it yourself is part of what this lab teaches.

Important Information

Warning

Important: How you will sign in to this lab

Microsoft requires Multifactor Authentication (MFA) on all accounts. To keep this lab simple and hassle-free, GetHandsOn.ai uses a Temporary Access Pass (TAP) instead.

Your TAP is a short-lived, one-time passcode that replaces your password and bypasses the MFA prompt entirely. No phone or authenticator app needed. You will find your TAP alongside your username in the Credentials panel on the right, and it stays visible throughout the session.

Read this before you start

In this hosted lab environment, there are restrictions on the resource names, credentials, and settings you can use. Follow the instructions carefully and use only the values provided. Most fields in this lab will be pre-populated for you - accept the defaults unless the lab explicitly tells you otherwise.

This is your first lab in the AI-103 course. You will not write any code in this lab. Your goal is to learn your way around the Microsoft Foundry portal so that every lab from Lab 2 onwards feels familiar.

Info

A note on pace. AI-103 assumes you're comfortable with the basics of Microsoft Foundry and ready to start building. If you've recently completed AI-901 on GetHandsOn.ai, some of the early labs will feel familiar - that's intentional, because the rest of the course builds directly on those foundations and not every learner arrives with the same background. If a section is review for you, move through it quickly. The technical depth increases as the course progresses.

What You Will Learn

✅ How to sign in to Microsoft Foundry using your Temporary Access Pass ✅ How to create a Foundry project - your workspace for everything you'll build ✅ A critical setting you must turn off before project creation will succeed ✅ The six top-bar menus and what each is for: Home, Discover, Build, Operate, Manage, Docs ✅ How to explore the model catalog, agent templates, and tool catalog ✅ How to deploy your first model and test it in the playground ✅ Where Knowledge bases, Memory, Evaluations, Guardrails and Tracing live under Build, and where Quota lives under Manage ✅ How to read your own quota, and why "Default settings" is the most expensive click in the portal ✅ How to copy your project credentials - the pattern every future lab uses

Key Terminology

Before you start clicking around, here are the words you will see repeatedly across this lab and the rest of the course. Skim now, refer back whenever a term feels fuzzy.

Info

Model - A trained AI system (for example gpt-5-mini) that takes some combination of text, image and audio as input and returns output. Which combination depends on the model, and the model card tells you. Models live in the catalog until you deploy them.

Deployment - A live, callable instance of a model inside your project. The same model can be deployed more than once with different settings.

Endpoint - The URL your code calls to talk to a deployment. Foundry gives you two: a Project endpoint (used by the Foundry SDK) and an Azure OpenAI endpoint (used by the OpenAI SDK).

Project - Your working space inside Foundry. It groups deployments, agents, knowledge bases, data, evaluations, and credentials together.

Token - The unit a model reads and writes. Roughly 4 characters of English text per token. Pricing and limits are counted in tokens.

Context window - The maximum number of tokens a model can consider in a single request (input plus output combined). gpt-5-mini has a 400k context window (272k input plus 128k output).

Agent - A model plus instructions, tools, and optional knowledge, packaged so it can carry out multi-step tasks.

Guardrail - A content-safety policy attached to a deployment that filters harmful input or output across categories like hate, sexual, violence, and self-harm.

Quota - The capacity limit (tokens per minute, requests per minute) your subscription has for a model in a region. If a lab fails with a quota error, this is what to check.

SKU - The pricing and performance tier you pick when deploying. Global Standard is the serverless default used across this course.

Welcome to Project Horizon

You are starting your first day as an AI Developer at Solara Travels - a boutique travel company with eight properties across Europe, Asia, and Africa. The CTO, Lena Voss, has launched a six-month initiative called Project Horizon to bring Azure AI into Solara's operations. Across the course you'll work with three Solara colleagues:

• Lena Voss - Chief Technology Officer (Phase 1 and Phase 4)
• Raj Menon - Product Manager (Phase 2)
• Amara Osei - Regional Director, West Africa & Europe (Phase 3)

For today there's no email from any of them. They're trusting you to find your bearings before tomorrow's standup. Let's get started.

Step 1: Sign In to Microsoft Foundry

Your lab environment is already running. You will see a Windows desktop with Microsoft Edge open.

Navigate to:

https://ai.azure.com/nextgen

Info

Why the /nextgen URL?

Microsoft Foundry has two views - a public marketing landing page at ai.azure.com and the actual workspace at ai.azure.com/nextgen. The /nextgen URL takes you straight to where you need to be and triggers sign-in if you're not already signed in.

Sign in with your Temporary Access Pass

1. Microsoft will show a Sign in screen.
2. Enter your username from the Credentials panel on the top (it looks like student-xxxxxx@GetHandsOn.ai).
3. Click Next.
4. Microsoft will show an Enter Temporary Access Pass screen. Paste the TAP from the Credentials panel.
5. Click Sign in.

Warning

Use only the credentials provided in the Credentials panel. Do not use your personal Microsoft account. Your lab environment is isolated and will be cleaned up automatically when the timer expires.