According to the Microsoft Azure AI Fundamentals (AI-900) Official Study Guide and the Microsoft Learn module “Explore Azure Machine Learning”, when you publish an inference pipeline (a deployed web service for real-time predictions) using Azure Machine Learning designer, you make the model accessible as a RESTful endpoint. Consumers—such as applications, scripts, or services—interact with this endpoint to submit data and receive predictions.

To securely access this deployed pipeline, two critical parameters are required:

REST endpoint (Option D):The REST endpoint is a URL automatically generated when the inference pipeline is deployed. It defines the network location where clients send HTTP POST requests containing input data (usually in JSON format). The endpoint routes these requests to the deployed model, which processes the data and returns prediction results. The REST endpoint acts as the primary access point for consuming the model’s inferencing capability programmatically.

Authentication key (Option C):The authentication key (or API key) is a security token provided by Azure to ensure that only authorized users or systems can access the endpoint. When invoking the REST service, the key must be included in the request header (typically as the value of the Authorization header). This mechanism enforces secure, authenticated access to the deployed model.

The other options are incorrect:

A. The model name is not required to consume the endpoint; it is used internally within the workspace.

B. The training endpoint is used for training pipelines, not for inference.

Therefore, according to Microsoft’s official AI-900 learning objectives and Azure Machine Learning documentation, when consuming a published inference pipeline, you must use both the REST endpoint (D) and the authentication key (C). These parameters ensure secure, controlled, and programmatic access to the deployed AI model for real-time predictions.

Model Explain Ability.

Most businesses run on trust and being able to open the ML “black box” helps build transparency and trust. In heavily regulated industries like healthcare and banking, it is critical to comply with regulations and best practices. One key aspect of this is understanding the relationship between input variables (features) and model output. Knowing both the magnitude and direction of the impact each feature (feature importance) has on the predicted value helps better understand and explain the model. With model explain ability, we enable you to understand feature importance as part of automated ML runs.

For evaluating a classification model, the appropriate metric from the options provided is the True Positive Rate (TPR), also known as Sensitivity or Recall. According to the Microsoft Azure AI Fundamentals (AI-900) official study guide and Microsoft Learn module “Evaluate model performance”, classification models are evaluated using metrics that measure how accurately the model predicts categorical outcomes such as “yes/no,” “spam/not spam,” or “approved/denied.”

The True Positive Rate measures the proportion of correctly identified positive cases out of all actual positive cases. Mathematically, it is expressed as:

True Positive Rate (Recall)=True PositivesTrue Positives + False Negatives\text{True Positive Rate (Recall)} = \frac{\text{True Positives}}{\text{True Positives + False Negatives}}True Positive Rate (Recall)=True Positives + False NegativesTrue Positives​

This metric is important when missing positive predictions carries a high cost, such as in medical diagnosis or fraud detection. Microsoft Learn highlights classification evaluation metrics such as accuracy, precision, recall, F1 score, and AUC (Area Under the Curve) as suitable for classification models.

The other options—Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Coefficient of Determination (R²)—are regression metrics used to evaluate models that predict numeric values rather than categories. For example, they apply to predicting house prices or temperatures, not yes/no decisions.

Therefore, the correct classification evaluation metric among the choices is A. True Positive Rate.

 “Generative AI enables software applications to generate new content, such as language dialogs and images.” — YES

This statement is true. According to the Microsoft Azure AI Fundamentals (AI-900) study guide and Microsoft Learn documentation, Generative AI refers to systems capable of creating new content such as text, audio, images, video, and code. Models like GPT, DALL·E, and Codex use deep learning to generate human-like responses, natural conversations, or creative media. This is a key differentiator between generative and discriminative AI — generative AI produces new data, while discriminative AI categorizes or analyzes existing data.

 “The difference between a large language model (LLM) and a small language model (SLM) is the number of variables in the model.” — YES

This statement is true. The primary distinction between an LLM and an SLM lies in the scale of parameters (variables) within the neural network. LLMs contain billions or even trillions of parameters, which enable them to capture complex linguistic patterns and perform broader tasks. SLMs have fewer parameters, making them faster but less capable of handling complex, context-rich tasks.

 “Generative AI is a type of supervised learning.” — NO

This statement is false. Generative AI models are typically trained using unsupervised or self-supervised learning methods. They learn by predicting missing or next elements in large text or image datasets rather than relying on labeled input-output pairs, which are used in supervised learning.

Full Detailed Explanation (250–300 words):

“You can fine-tune some Azure OpenAI models by using your own data.” – YESThis statement is true. Azure OpenAI allows customers to fine-tune certain models like GPT-3, GPT-3.5, and some embedding models with their own data. Fine-tuning customizes a model to perform better on specific tasks or match a company’s domain terminology, tone, or context. According to Microsoft Learn’s AI-900 and Azure OpenAI documentation, fine-tuning is supported for approved use cases while maintaining Microsoft’s Responsible AI oversight and compliance process.

“Pretrained generative AI models are a component of Azure OpenAI.” – YESThis statement is also true. Azure OpenAI provides access to pretrained large language and generative AI models such as GPT-3.5, GPT-4, Codex, and DALL·E. These models are pretrained on vast datasets and made available via APIs, allowing developers to generate text, code, and images without needing to train their own models. This is a core feature of Azure OpenAI’s service offering.

“To build a solution that complies with Microsoft responsible AI principles, you must build and train your own model.” – NOThis statement is false. Compliance with Microsoft Responsible AI principles (Fairness, Reliability & Safety, Privacy & Security, Inclusiveness, Transparency, Accountability) does not require building custom models. Prebuilt Azure AI and OpenAI services already align with Responsible AI standards. Developers simply need to use these services responsibly, applying governance and ethical design practices.

According to the Microsoft Azure AI Fundamentals (AI-900) official study guide and the Microsoft Learn module “Prepare data for machine learning”, feature engineering refers to the process of transforming raw data into meaningful features that can be effectively used by machine learning algorithms. This includes steps such as scaling, normalization, encoding categorical variables, handling missing values, and creating new features derived from existing ones.

The question states:

“Ensuring that the numeric variables in training data are on a similar scale.”

This directly describes a data normalization or standardization step, which is a core component of feature engineering. The purpose of scaling numeric variables is to ensure that all features contribute equally to the model’s learning process. Without normalization, features with large numeric ranges (such as “income in dollars”) could dominate smaller-scale features (like “age in years”), leading to biased model performance.

In Azure Machine Learning, this is typically done using the Normalize Data module or transformations in the data preparation stage. Microsoft Learn explains that normalization and feature scaling are applied before model training to ensure that gradient-based algorithms (such as regression or neural networks) converge more efficiently and produce more accurate results.

The other options are not correct:

Data ingestion refers to collecting and importing data into a system.

Feature selection involves choosing the most relevant features, not scaling them.

Model training is the phase where the algorithm learns patterns from the processed data, which occurs after feature engineering.

Therefore, ensuring that numeric variables are on a similar scale is a step in Feature Engineering.

The correct completion of the sentence is:

“The interactive answering of questions entered by a user as part of an application is an example of natural language processing.”

According to the Microsoft Azure AI Fundamentals (AI-900) official study materials, Natural Language Processing (NLP) is a branch of Artificial Intelligence that focuses on enabling computers to understand, interpret, and respond to human language in a way that is both meaningful and useful. It is one of the key AI workloads described in the “Describe features of common AI workloads” module on Microsoft Learn.

When a user types a question into an application and the system responds interactively — such as in a chatbot, Q & A system, or virtual assistant — this process requires language understanding. NLP allows the system to process the input text, determine user intent, extract relevant entities, and generate an appropriate response. This is the foundational capability behind services such as Azure Cognitive Service for Language, Language Understanding (LUIS), and QnA Maker (now integrated as Question Answering in the Language service).

Microsoft’s study guide explains that NLP workloads include the following key scenarios:

Language understanding: Determining intent and context from text or speech.

Text analytics: Extracting meaning, key phrases, sentiment, or named entities.

Conversational AI: Powering bots and virtual agents to interact using natural language.These systems rely on NLP models to analyze user inputs and respond accordingly.

In contrast:

Anomaly detection identifies data irregularities.

Computer vision analyzes images or video.

Forecasting predicts future values based on historical data.

Therefore, based on the AI-900 official materials, the interactive answering of user questions through an application clearly falls under Natural Language Processing (NLP).

According to the Microsoft Azure AI Fundamentals (AI-900) curriculum and Microsoft Learn’s “Explore Automated Machine Learning in Azure Machine Learning” module, Automated ML (AutoML) is the Azure Machine Learning capability that allows users to quickly build, train, and deploy predictive models with minimal or no coding experience.

Automated ML automatically performs tasks that would normally require expert data science knowledge, such as:

Selecting appropriate algorithms (e.g., decision trees, logistic regression, random forests)

Performing hyperparameter tuning to optimize model accuracy

Handling missing data and feature scaling automatically

Generating performance metrics and best model recommendations

This feature is especially useful for business analysts, developers, or beginners who want to leverage machine learning for predictions (like sales forecasting, churn analysis, or demand prediction) without having to write complex Python code.

Other options explained:

A. ML pipelines automate and organize workflows for model training and deployment but still require pre-built models.

B. Copilot is an AI-powered assistant embedded in Microsoft tools for productivity, not a model training feature.

C. DALL-E is an image generation model under Azure OpenAI, not a predictive modeling tool.

Thus, per official Microsoft Learn content, Automated Machine Learning is the correct capability to quickly build and deploy predictive models with minimal coding.

According to the Microsoft Azure AI Fundamentals (AI-900) official study guide and Microsoft Learn module “Identify features of Computer Vision workloads on Azure”, the Custom Vision service is a part of Azure Cognitive Services that allows users to build, train, and deploy custom image classification and object detection models. It is primarily designed for still-image analysis, not video processing.

“The Custom Vision service can be used to detect objects in an image.” – Yes.This is correct. The Custom Vision service supports two major model types: classification (categorizing entire images) and object detection (identifying and locating multiple objects within a single image). In object detection mode, the model outputs both the object’s category and its position in the image using bounding boxes. This capability is emphasized in the AI-900 curriculum as an example of applying computer vision to real-world scenarios, such as identifying products on shelves or detecting equipment parts in manufacturing.

“The Custom Vision service requires that you provide your own data to train the model.” – Yes.This statement is also true. Unlike prebuilt computer vision models, Custom Vision is a trainable model that requires users to upload their own labeled images to create a domain-specific AI model. The model’s accuracy depends on the quality and quantity of this user-provided data. The AI-900 study materials explain that Custom Vision is used when prebuilt models do not meet specific needs, enabling businesses to train models tailored to unique image sets.

“The Custom Vision service can be used to analyze video files.” – No.This is incorrect. Custom Vision is limited to image-based analysis. To analyze video content (detecting objects or motion in moving frames), Azure provides Video Indexer, which is a separate service designed for extracting insights from video files, including speech, objects, faces, and emotions.

Confidence.

According to the Microsoft Azure AI Fundamentals (AI-900) Official Study Guide and the Microsoft Learn module “Explore computer vision in Microsoft Azure,” the confidence score represents the calculated probability that a model’s prediction is correct. In image classification, when an AI model analyzes an image and assigns it to a specific category, it also produces a confidence value—a numerical probability (usually between 0 and 1) indicating how certain the model is about its prediction.

For example, if an image classification model identifies an image as a “cat” with a confidence of 0.92, it means the model is 92% certain that the image depicts a cat. The confidence value helps developers and users understand the model’s certainty level about its classification output.

Microsoft Learn emphasizes that in Azure Cognitive Services—such as the Custom Vision Service—each prediction result includes both the predicted label (class) and a confidence score. These confidence scores are essential for evaluating model performance and determining thresholds for automated decisions (e.g., accepting predictions only above a 0.8 probability).

Let’s evaluate the other options:

Accuracy: This is an overall performance metric measuring the percentage of correct predictions across the dataset, not a probability for a single prediction.

Root Mean Square Error (RMSE): This is a metric for regression models, not classification tasks. It measures average error magnitude between predicted and actual values.

Sentiment: This is a type of prediction (positive, negative, neutral) in text analysis, not a probability metric.

Therefore, based on Microsoft’s AI-900 training materials and Azure Cognitive Services documentation, the calculated probability of a correct image classification is called Confidence, which expresses how sure the model is about its prediction for a specific input.