AI-900 Practice Exam 51 — 20 Questions

Question 1

A developer uses Azure OpenAI Service to generate code. They provide a few examples of function definitions and their corresponding descriptions, then ask the model to write a new function based on a new description. Which technique is the developer using?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Prompt engineering with few-shot learning. The developer is using prompt engineering with few-shot learning, a technique where a small set of input-output examples (here, function definitions and descriptions) is included in the prompt to guide the model's behavior without modifying its weights. This leverages the model's in-context learning ability to generalize from the provided examples and generate a new function for a new description.

Answer

Fine-tuning the model with the examples

Answer

Training a custom model from scratch

Answer

Using reinforcement learning from human feedback

Question 2

What is 'image recognition' as a core AI workload capability?

Accepted Answer

AI capabilities for identifying and classifying visual content including objects, faces, and text. Image recognition is a core AI workload capability that enables systems to identify and classify visual content such as objects, faces, and text within images. This is typically achieved using convolutional neural networks (CNNs) trained on large labeled datasets to extract features and make predictions. It is distinct from image generation or storage, focusing on understanding existing visual data.

Answer

Software that controls physical hardware cameras and their settings

Answer

Generating new images from text descriptions using AI

Answer

Storing and retrieving images from a database using unique identifiers

Question 3

What is 'dense captioning' in Azure AI Vision v4.0?

Accepted Answer

Generating multiple region-specific captions each with a bounding box for different image areas. Dense captioning in Azure AI Vision v4.0 goes beyond describing the entire image; it identifies multiple distinct regions within the image and generates a separate caption for each region, along with a bounding box that pinpoints its location. This allows for granular understanding of complex scenes, such as recognizing 'a dog on a couch' and 'a lamp on a table' as separate, localized descriptions.

Answer

Generating a very long and detailed caption for the entire image

Answer

Adding caption text overlaid on top of the image like movie subtitles

Answer

Captions that include technical details like camera settings and lighting conditions

Question 4

What is 'Azure Machine Learning datasets' and why are they important?

Accepted Answer

Versioned, registered data references enabling reproducibility, sharing, and lineage tracking in Azure ML. Azure Machine Learning datasets are versioned, registered data references that encapsulate metadata such as location, schema, and creation time, enabling reproducibility, sharing, and lineage tracking across experiments. They do not store the raw data files themselves but provide a pointer to the data source (e.g., Azure Blob Storage, Azure Data Lake), ensuring that every training run uses the exact same data snapshot, which is critical for auditability and collaboration.

Answer

The raw data files stored in Azure Blob Storage before any processing

Answer

Synthetic datasets automatically generated by Azure ML to supplement small training sets

Answer

Pre-labelled benchmark datasets provided by Microsoft for testing Azure ML models

Question 5

What is 'neural architecture search' (NAS) and how does it relate to AutoML?

Accepted Answer

Automating the discovery of optimal neural network architectures using computational search. Neural Architecture Search (NAS) is an automated process that uses computational search methods—such as reinforcement learning, evolutionary algorithms, or gradient-based optimization—to discover optimal neural network architectures for a given task. It is a key component of AutoML because AutoML aims to automate the entire machine learning pipeline, including model selection and hyperparameter tuning, and NAS specifically automates the design of the neural network topology itself.

Answer

Searching the web for neural network architectures published in research papers

Answer

Querying a database of pre-built neural networks to find the closest match for a task

Answer

A legal search process for patenting new AI model architectures

Question 6

What is 'multi-label text classification' vs 'single-label' in Azure AI Language?

Accepted Answer

Single-label assigns exactly one category; multi-label allows multiple categories per document. In Azure AI Language, single-label text classification assigns exactly one category to each document, while multi-label classification allows a document to be assigned multiple categories simultaneously. This distinction is fundamental to how the classification models are trained and how predictions are structured, with multi-label using a separate binary classifier per label rather than a single softmax output.

Answer

Single-label classifies one word; multi-label classifies entire sentences

Answer

Multi-label is more accurate because it considers more information per document

Answer

Single-label requires more training data than multi-label classification

Question 7

What is 'clustering' in unsupervised machine learning?

Accepted Answer

Grouping similar data points together without predefined labels based on natural patterns. Clustering is an unsupervised learning technique that automatically groups data points based on inherent similarities or patterns in the data, without requiring any pre-existing labels. The algorithm identifies natural structures, such as distance or density relationships, to form clusters. In Azure Machine Learning, clustering is commonly implemented using algorithms like K-Means or DBSCAN for tasks such as customer segmentation or anomaly detection.

Answer

Classifying data points into predefined categories using labelled training examples

Answer

Grouping Azure compute resources together for distributed training jobs

Answer

Organising model training runs into logical groups for experiment tracking

Question 8

A global e-commerce company receives customer support emails in over 30 languages. They want to automatically route each email to the correct department (Billing, Technical Support, or Returns) and also extract the order number and the product name mentioned in the email. The solution must be able to identify the language of each email before further processing. Which combination of Azure AI Language features should they use?

Accepted Answer

Language Detection, Custom Text Classification, and Entity Recognition. The scenario requires three distinct capabilities: Language Detection to identify the email's language (a prerequisite for further processing), Custom Text Classification to route emails to the correct department (Billing, Technical Support, or Returns), and Entity Recognition to extract the order number and product name. Azure AI Language provides these as separate, combinable features that directly map to the stated requirements.

Answer

Language Detection, Key Phrase Extraction, and Sentiment Analysis

Answer

Entity Recognition, Sentiment Analysis, and Key Phrase Extraction

Answer

Custom Text Classification, Key Phrase Extraction, and Sentiment Analysis

Question 9

What is 'Azure OpenAI's fine-tuning' feature and what data format does it require?

Accepted Answer

Training a base model on domain-specific JSONL conversation examples to adapt its behaviour. Azure OpenAI's fine-tuning feature allows you to take a pre-trained base model (such as GPT-3.5 or GPT-4) and further train it on your own domain-specific dataset to improve its performance on particular tasks. The required data format is JSONL (JSON Lines), where each line contains a conversation example structured with a 'messages' array that includes 'role' (system, user, assistant) and 'content' fields. This process adapts the model's behavior without altering its core architecture, making it more accurate for specialized use cases like customer support or legal document analysis.

Answer

A feature for adjusting model parameters in real time based on user feedback during deployment

Answer

A no-code interface for adjusting temperature and top_p settings without writing code

Answer

Restricting the model to only generate responses related to topics in your training data

Question 10

What is 'evaluation' of generative AI models in Azure AI Foundry?

Accepted Answer

Systematically measuring a generative AI application's quality (groundedness, relevance) and safety metrics. In Azure AI Foundry, evaluation refers to the systematic measurement of a generative AI application's quality and safety using predefined metrics such as groundedness (factual alignment with source data), relevance, and safety (e.g., content filtering). This process is distinct from ad-hoc user feedback or training diagnostics, as it provides structured, repeatable assessments to validate model behavior before deployment.

Answer

The process of assessing job candidates using AI-powered assessments

Answer

Having users rate the AI's responses with thumbs up or thumbs down during beta testing

Answer

Running model training and measuring loss curves to determine when to stop training

Question 11

What is 'multi-language support' in Azure AI Language and why does it matter?

Accepted Answer

NLP capabilities (sentiment, NER, etc.) that work across 100+ human languages for global applications. Azure AI Language provides pre-built NLP capabilities—such as sentiment analysis, named entity recognition (NER), key phrase extraction, and language detection—that are trained to work across more than 100 human languages. This multi-language support is critical for global applications that need to process user input in diverse languages without requiring separate models or custom training for each language.

Answer

The ability to write Azure AI Language SDK code in multiple programming languages

Answer

Translating all NLP model outputs into the user's preferred language automatically

Answer

Combining multiple NLP models that each specialise in a different language

Question 12

What is 'AI at the edge' and why would you deploy an AI model to an edge device?

Accepted Answer

Running AI inference locally on devices for low latency, offline capability, and data privacy. B is correct because 'AI at the edge' refers to running AI inference locally on edge devices (e.g., IoT sensors, cameras, or local servers) rather than in the cloud. This approach provides low latency by processing data immediately without network round-trips, enables offline capability when connectivity is intermittent, and enhances data privacy by keeping sensitive data on the device. It is a core AI workload consideration for scenarios like real-time video analytics or industrial predictive maintenance.

Answer

Using AI to analyse data collected near the geographic borders of a country

Answer

Using AI to detect adversarial attacks at the network perimeter

Answer

Deploying AI to the most remote Azure region for disaster recovery

Question 13

What is a 'system message' (system prompt) in Azure OpenAI chat models?

Accepted Answer

A developer-set instruction that defines the model's role, persona, and behavioural constraints. A system message (system prompt) in Azure OpenAI chat models is a developer-defined instruction that sets the model's role, persona, and behavioral constraints. This prompt is sent as part of the conversation context to guide the model's responses, ensuring it adheres to specific guidelines, tone, or safety rules. It is not an error notification, automated confirmation, or user input.

Answer

An error notification sent by Azure when the OpenAI service is unavailable

Answer

Automated messages the model sends to confirm it received the user's input

Answer

The first message a user sends to start a new conversation session

Question 14

A marketing team uses Azure OpenAI Service to generate tagline options for a new product. They notice that the model often generates very similar taglines for the same prompt, lacking creativity. To increase the diversity and variety of the output, which parameter should they increase?

Accepted Answer

Temperature. Increasing the temperature parameter makes the model's output more random and diverse by scaling the probability distribution over possible next tokens. A higher temperature (e.g., 0.9) flattens the distribution, giving lower-probability tokens a better chance to be selected, which directly addresses the lack of creativity and variety in the generated taglines.

Answer

Top P

Answer

Frequency penalty

Answer

Max tokens

Question 15

What is 'Azure OpenAI deployment' and how does it differ from a 'model'?

Accepted Answer

A model is the underlying AI; a deployment is a named, quota-allocated instance your application calls. In Azure OpenAI, a 'model' refers to the underlying AI algorithm (e.g., GPT-4, GPT-3.5-Turbo) that defines the capabilities and behavior of the generative AI. A 'deployment' is a specific, named instance of that model provisioned within an Azure OpenAI resource, with its own endpoint, quota (tokens per minute), and configuration (e.g., content filter settings). This separation allows you to manage capacity and access for different applications or use cases independently, even when using the same base model.

Answer

A model is the purchased licence; a deployment is the technical installation

Answer

A deployment is always faster than a model because it uses optimised serving infrastructure

Answer

Models are available globally; deployments are restricted to specific Azure regions

Question 16

What is recall (sensitivity) in the context of binary classification model evaluation?

Accepted Answer

The proportion of actual positives that the model correctly identified. Recall (sensitivity) measures the proportion of actual positive cases that the model correctly identifies. In binary classification, it answers: 'Of all the truly positive instances, how many did the model catch?' This is critical in scenarios where missing a positive (false negative) is costly, such as disease screening or fraud detection.

Answer

The proportion of positive predictions that are actually correct

Answer

The overall proportion of predictions that match the actual labels

Answer

How quickly the model can be updated with new training data

Question 17

What is 'MLOps' and how does it relate to AI workloads on Azure?

Accepted Answer

Applying DevOps practices (automation, CI/CD, monitoring) to the machine learning lifecycle. MLOps (Machine Learning Operations) is the application of DevOps principles—such as automation, continuous integration/continuous deployment (CI/CD), and monitoring—to the machine learning lifecycle. On Azure, MLOps is implemented through services like Azure Machine Learning, which provides pipelines, model registries, and automated retraining to manage the end-to-end ML workflow from data preparation to deployment and monitoring.

Answer

Operational procedures for Microsoft 365 mail system administration

Answer

A certification program for ML engineers working with Azure

Answer

The process of optimising ML model inference speed for production deployment

Question 18

A law firm needs to automatically detect and redact sensitive information such as names, addresses, and social security numbers from legal documents. Which Azure AI Language feature can detect these entities without custom training?

Accepted Answer

PII Detection. PII Detection is the correct Azure AI Language feature because it is specifically designed to identify and redact sensitive personal information such as names, addresses, and social security numbers from text without requiring any custom training. This pre-built capability uses machine learning models to detect categories of personally identifiable information (PII) out of the box, making it ideal for compliance scenarios like legal document processing.

Answer

Sentiment Analysis

Answer

Key Phrase Extraction

Answer

Language Detection

Question 19

A company wants to automatically categorize support tickets into categories such as 'Billing', 'Technical Issue', and 'Account Management'. They have a set of 1,000 labeled tickets that they can use to train a model. Which Azure AI Language feature should they use?

Accepted Answer

Custom text classification. Custom text classification is the correct choice because it allows you to train a model on your own labeled dataset (1,000 tickets) to categorize text into custom-defined classes like 'Billing', 'Technical Issue', and 'Account Management'. This feature is specifically designed for scenarios where you need to classify documents into user-defined categories, unlike pre-built classification options.

Answer

Key phrase extraction

Answer

Sentiment analysis

Answer

Language detection

Question 20

A retail company deploys an AI system that analyzes customer purchase history to personalize product recommendations. Without informing customers, the system also uses their names, addresses, and phone numbers to create detailed profiles. A customer advocacy group raises concerns about this practice. Which Microsoft responsible AI principle is most directly violated?

Accepted Answer

Privacy and security. (Privacy and security) because the AI system collects and uses customers' personally identifiable information (PII) such as names, addresses, and phone numbers without their knowledge or consent. This directly violates the Microsoft responsible AI principle of Privacy and security, which mandates that data collection and usage must be transparent, consensual, and protected against unauthorized access. The scenario describes a clear breach of data governance and user consent, which is the core of this principle.

Answer

Fairness

Answer

Reliability and safety

Answer

Transparency