Skip to main content

Practice exam · Azure

Microsoft Azure AI Engineer Associate (AI-102) practice exam

50
questions
120
minutes
70%
to pass the practice exam
60
questions in the bank

Short answer

The Microsoft Azure AI Engineer Associate (AI-102) practice exam has 50 questions in 120 minutes, with a 70% passing mark, as configured here. The fastest way to know whether you are ready is to sit a full exam on the clock and read your score by domain, not only the total.

Domains covered

  • Implement Computer Vision Solutions
  • Implement Conversational AI Solutions
  • Implement Generative AI Solutions
  • Implement Knowledge Mining and Document Intelligence
  • Implement Natural Language Processing Solutions
  • Plan and Manage an Azure AI Solution

Sample questions, with explanations

Three questions from the bank, open to read.

Question 1 · Implement Knowledge Mining and Document Intelligence

You must extract data from a set of vendor invoices that all follow different, unpredictable layouts and contain a mix of printed and handwritten fields. You have 120 labeled sample documents across the layout variations. Which Azure AI Document Intelligence approach gives the most reliable field extraction?

  1. Train a custom template model on the labeled samples
  2. Train a custom template model per layout, then compose them so no neural training is needed
  3. Train a custom neural model on the labeled samples ✓ (correct)
  4. Use the prebuilt-layout model and parse fields from the returned bounding boxes yourself
Why this is correct

Custom neural models are designed for structured-to-unstructured and mixed-type documents whose layouts vary, and they handle variation in field positions and handwriting far better than template matching — making them the right choice for invoices with unpredictable layouts. A single custom template model is the wrong fit because template models rely on consistent visual positioning (layout cues at fixed coordinates); they degrade badly when layouts vary widely. Using prebuilt-layout and parsing bounding boxes yourself extracts structure (tables, lines, selection marks) but not the semantic key-value fields you want, pushing all the field-mapping logic onto you with no learned generalization. Composing many per-layout template models still inherits the template fragility for each unseen layout variation, and composition only routes inputs to the best-matching sub-model — it does not add the generalization that the variable, handwriting-heavy invoices require.

Question 2 · Implement Knowledge Mining and Document Intelligence

A developer wants one Azure AI Document Intelligence endpoint call to automatically route an incoming document to the correct custom extraction model depending on whether it is a 'PurchaseOrder' or a 'Receipt', each of which has its own trained model. Which capability should they use?

  1. Call the prebuilt-read model first and branch on the detected language
  2. Enable incremental training so the receipt model learns purchase-order fields over time
  3. Create a composed model from the two custom models and analyze against the composed model ID ✓ (correct)
  4. Train a single custom neural model that includes both document types' fields in one schema
Why this is correct

A composed model groups multiple standalone custom models under one model ID; at analysis time the service classifies the input and routes it to the best-matching sub-model, returning that model's fields — exactly the auto-routing requested. Merging both document types into one custom neural schema is possible but conflates two distinct field sets into a single model, which is harder to maintain, can blur field semantics, and is not the documented routing mechanism. Calling prebuilt-read and branching on detected language is wrong because prebuilt-read does OCR/text and language detection, not document-type classification or custom-field routing. 'Incremental training so the receipt model learns purchase-order fields' is a misconception — Document Intelligence custom models are retrained from a labeled set; there is no online incremental-learning mode that absorbs a second document type into an existing model at inference time.

Question 3 · Plan and Manage an Azure AI Solution

An enterprise standard requires that all secrets — including any Azure AI Services keys still in use during a phased migration to Entra ID — be stored centrally and never appear in app settings or source control. The current Function App reads the Azure AI key from an environment variable. Which plan best aligns with the standard while preparing for keyless auth?

  1. Store the key as an Azure Key Vault secret and grant the Function App's managed identity 'Key Vault Secrets User' so it retrieves the secret at runtime; concurrently move toward 'Cognitive Services User' RBAC to drop keys entirely. ✓ (correct)
  2. Move the key into the Function App's application settings and mark it as a 'slot setting' so it is encrypted at rest and excluded from source control automatically.
  3. Embed the key in the deployment package and rotate it weekly via a release pipeline variable, which keeps it out of source control while remaining readable by the app.
  4. Grant the Function App's managed identity the 'Key Vault Administrator' role so it has full control of the vault and can both read and rotate the AI key as needed.
Why this is correct

The central-secret-store pattern is Azure Key Vault accessed via managed identity. Storing the AI key as a Key Vault secret and granting the Function App's identity the least-privilege 'Key Vault Secrets User' (read secrets only) keeps the key out of app settings and source control, and the parallel move to 'Cognitive Services User' RBAC achieves the keyless end state. Wrong: 'app settings marked as slot setting' misunderstands slot settings — they control swap behavior, not secret governance; the value still lives in app configuration, violating the standard. Wrong: 'embed in the deployment package' literally places the secret in build artifacts, the opposite of central, non-source-controlled storage. Wrong: 'grant Key Vault Administrator' violates least privilege — the app only needs to READ a secret; Administrator grants destructive control and is an over-permission governance mistake.

Take the full practice exam

  • 50 questions drawn from the bank, with a 120-minute timer
  • Score by domain and a review of every question
  • An AI tutor to talk through each question you miss
  • What you miss feeds your study plan
Start the practice exam

Independent, unofficial practice exam. Azure and the exam name are trademarks of their owners; UpSkillScore is not affiliated with them. Format and passing mark are the exam's configuration in UpSkillScore, not an official statement.

Frequently asked questions

How many questions does the Microsoft Azure AI Engineer Associate (AI-102) practice exam have?

It has 50 questions and a 120-minute limit. The passing mark of the practice exam is 70%.

Is this the official Microsoft Azure AI Engineer Associate (AI-102) exam?

No. It is an independent practice exam built to match the structure of the real one. It does not replace the official exam.

Do I need an account to take it?

You can read the structure and sample questions without an account. Taking the full timed exam, scoring by domain and the AI tutor need one.