Free · No sign-in required
15 real questions from our 65-question AI-901 mock, spanning both exam domains. Answer them, then get an exam-weighted score against the 70% pass line, a readiness verdict, and a per-lesson plan for your weakest areas — no account needed.
Free sample · 15 of 65 questions
0 / 15 answered
A factory streams sensor readings and wants to flag readings that fall outside normal operating ranges. Which AI workload best fits?
You need to predict the selling price of a house from its size and location. Which type of machine learning model is appropriate?
A model scores 99% on training data but performs poorly on new data. This is most likely:
Producing new marketing copy or images from a text prompt is which workload?
You need not only to know an image contains cars, but also where each car is located with a bounding box. Which workload do you need?
A hiring model approves men at a far higher rate than equally qualified women. Which Responsible AI principle is most directly violated?
For high-impact decisions such as medical triage, the recommended governance practice is to:
In Azure AI Foundry, the workspace that groups your models, data connections, and deployments for an app is called a:
Large language models break text into units for processing and billing. These units are called:
You want a chatbot to answer using your company's own PDF documents rather than only the model's general knowledge. The recommended approach is:
Which Azure AI Language feature determines whether text expresses positive, negative, or neutral opinion?
Generating a written transcript from a recorded meeting audio file uses which capability?
Which Azure AI Vision capability returns descriptive captions and tags for an image?
To extract totals, dates, and line items from invoices using a ready-made model, you would use:
Which Azure service indexes large content sets so users (and RAG apps) can query them, forming the basis of knowledge mining?