Pass the Microsoft Certified: Azure AI Fundamentals AI-900 Questions and answers with Dumpstech

The correct answers are B. assigning the color pixels in an image to object names and C. describing the contents of an image.

According to the Microsoft Azure AI Fundamentals (AI-900) official study guide and Microsoft Learn module “Describe features of computer vision workloads on Azure,” computer vision is a branch of AI that enables systems to analyze, interpret, and understand visual data from images and videos. It allows machines to identify objects, people, text, and even describe scenes automatically.

Option B: Assigning color pixels in an image to object names represents image classification or object detection, which are key computer vision workloads. In these tasks, AI analyzes pixel patterns to determine which pixels correspond to specific objects (for example, classifying pixels as “car,” “tree,” or “road”).

Option C: Describing the contents of an image corresponds to image captioning, another computer vision workload. It involves using AI models trained to generate natural language descriptions of what is visible in an image, such as “A group of people sitting at a dining table.” Azure’s Computer Vision service provides this functionality through its “Describe Image” API.

Incorrect options:

A. Creating photorealistic images involves generative AI and 3D modeling, not traditional computer vision.

D. Detecting inconsistencies and anomalies in a data stream relates to anomaly detection, not computer vision.

E. Creating visual representations of numerical data involves data visualization, not AI-driven image analysis.

In Conversational Language Understanding (CLU), a core service within Azure AI Language, intents represent the goals or purposes behind user utterances (for example, “Track my order” or “Cancel my subscription”). However, in real-world scenarios, users often provide inputs that do not match any defined intent. To handle such cases gracefully, Microsoft recommends including a “None” intent that captures out-of-scope utterances — text that doesn’t belong to any other intent in your model.

According to the Microsoft Learn module: “Build a Conversational Language Understanding app”, the None intent serves as a catch-all or fallback category for utterances that the model should ignore or respond to with a default message (e.g., “I’m sorry, I don’t understand that.”). By training the model with multiple examples of irrelevant or unrelated utterances in this intent, you improve its ability to distinguish between valid and invalid user inputs.

The other options are incorrect:

A. Export the model: Exporting only saves or transfers the model; it does not influence how the model detects irrelevant utterances.

B. Create a new model: A new model would not inherently solve out-of-scope detection unless properly trained with a None intent.

D. Create a prebuilt task entity: Entities identify specific data (like dates or products) within valid intents, not irrelevant utterances.

Thus, the correct approach to ensure that your CLU model can detect utterances outside its intended scope is to add examples of unrelated or off-topic phrases to the None intent. This improves classification accuracy and prevents incorrect intent matches.

✅ Correct Answer: C. Add utterances to the None intent

According to the Microsoft Azure AI Fundamentals (AI-900) official study guide and Microsoft Learn module “Describe features of computer vision workloads on Azure”, Optical Character Recognition (OCR) is a core computer vision workload that enables AI systems to detect and extract text from images or scanned documents.

In this scenario, the goal is to identify street names from street signs in photographs. Since the text is embedded within images, OCR is the correct technology to use. OCR works by analyzing the visual patterns of letters, numbers, and symbols, then converting them into machine-readable text. Azure’s Computer Vision API and Azure AI Vision Service provide OCR capabilities that can extract printed or handwritten text from pictures, documents, and even real-time camera feeds.

Let’s analyze the other options:

A. Object detection: Identifies and locates objects (like cars, people, or street signs) but not the text written on them.

C. Image classification: Classifies an entire image into categories (e.g., “street scene” or “traffic sign”) but doesn’t extract text content.

D. Facial recognition: Identifies or verifies people by analyzing facial features, unrelated to text extraction.

Therefore, identifying street names on street signs is a text extraction problem, making Optical Character Recognition (OCR) the most accurate and verified answer per Microsoft Learn content.

In Azure OpenAI Service, model behavior during text or chat generation is controlled by several parameters, such as temperature, max tokens, top_p, presence penalty, and frequency penalty. According to Microsoft Learn’s documentation for Azure OpenAI GPT models, the presence penalty influences how likely the model is to introduce new or diverse tokens in its responses.

Specifically, the presence penalty discourages the model from repeating previously used tokens, encouraging it to explore new topics or ideas instead of sticking to existing ones. Increasing the presence penalty value typically results in more diverse and creative outputs, while lowering it makes responses more repetitive or focused.

Option analysis:

A. Max response (Max tokens): Controls the maximum length of the generated response, not its diversity.

B. Past messages included: Defines how much chat history the model considers for context; it doesn’t affect diversity directly.

C. Presence penalty: Encourages novelty and introduces new tokens—this is correct for increasing response variety.

D. Stop sequence: Specifies a sequence of characters or tokens where the model should stop generating output.

According to the Microsoft Azure AI Fundamentals (AI-900) official study guide and Microsoft Learn module “Identify features of regression machine learning”, regression is a type of supervised machine learning used when the target variable (the value you want to predict) is a continuous numeric value.

In this scenario, the task is to predict how many hours of overtime a delivery person will work based on the number of orders received. Both the input (number of orders) and the output (hours of overtime) are numeric variables. Since the goal is to estimate a quantitative value rather than categorize or group data, this is a classic example of a regression problem.

Regression models analyze the relationship between variables to make numerical predictions. For example, the model might learn that each additional 20 orders increases overtime by about two hours. Common algorithms used for regression include linear regression, decision tree regression, and boosted regression models. These models produce outputs such as “expected overtime = 5.6 hours,” which are continuous numeric results.

To contrast with the other options:

Classification is used for predicting categories or labels, such as “overtime required” vs. “no overtime,” or “high-risk” vs. “low-risk.” It deals with discrete outputs rather than continuous numbers.

Clustering is an unsupervised learning approach used to group similar data points based on shared characteristics, such as grouping delivery staff by performance patterns or customer types.

As emphasized in Microsoft’s Responsible AI and Machine Learning Fundamentals learning paths, regression models are ideal for numeric forecasting problems such as predicting sales, revenue, demand, or working hours.

Therefore, the correct answer is: Regression.

B: Azure ' s Computer Vision service provides developers with access to advanced algorithms that process images and return information based on the visual features you ' re interested in. For example, Computer Vision can determine whether an image contains adult content, find specific brands or objects, or find human faces.

C: Computer Vision includes Optical Character Recognition (OCR) capabilities. You can use the new Read API to extract printed and handwritten text from images and documents.

The correct answers are C. the Azure OpenAI GPT-4 model and D. the GitHub Copilot service.

According to the Microsoft Azure AI Fundamentals (AI-900) curriculum and Microsoft Learn documentation on Azure OpenAI and GitHub Copilot, both GPT-4 and GitHub Copilot can be used to analyze and generate explanations for code functionality, as well as produce or refine code comments.

Azure OpenAI GPT-4 model (C):The GPT-4 model is a large language model (LLM) developed by OpenAI and available through the Azure OpenAI Service. It is trained on vast amounts of text, including programming languages, documentation, and natural language instructions. This enables it to interpret source code, explain what it does, suggest optimizations, and automatically generate detailed code comments. When prompted with code snippets, GPT-4 can provide structured natural language explanations describing the logic and intent of the code. In enterprise scenarios, developers use Azure OpenAI GPT models for code understanding, review automation, and documentation generation.

GitHub Copilot service (D):GitHub Copilot, powered by OpenAI Codex, is an AI coding assistant integrated into IDEs such as Visual Studio Code. It can analyze code context and generate inline comments and explanations in real time. GitHub Copilot understands the syntax and intent of numerous programming languages and provides intelligent suggestions or explanations directly in the developer’s environment.

The other options are not suitable:

A. DALL-E is a generative image model for creating visual content, not text or code analysis.

B. Whisper is an automatic speech recognition (ASR) model used for converting speech to text, unrelated to code interpretation.

Therefore, based on the official Azure AI and GitHub documentation, the correct and verified answers are C. Azure OpenAI GPT-4 model and D. GitHub Copilot service.

In the Microsoft Azure AI Fundamentals (AI-900) curriculum, computer vision capabilities refer to artificial intelligence systems that can analyze and interpret visual content such as images and videos. The Azure AI Vision and Face API services provide pretrained models for detecting, recognizing, and analyzing visual information, enabling developers to build intelligent applications that understand what they " see. "

When asked how computer vision capabilities can be deployed, the correct answer is to integrate a face detection feature into an app. This aligns with Microsoft Learn’s module “Describe features of computer vision workloads,” which explains that computer vision can identify objects, classify images, detect faces, and extract text (OCR). The Face API, a part of Azure AI Vision, specifically provides face detection, verification, and emotion recognition capabilities.

Integrating these services into an application allows it to perform actions such as:

Detecting human faces in photos or video streams.

Recognizing facial attributes like age, emotion, or head pose.

Enabling secure authentication based on face recognition.

The other options are incorrect because they relate to different AI workloads:

Develop a text-based chatbot for a website: This falls under Conversational AI, implemented with Azure Bot Service or Conversational Language Understanding (CLU).

Identify anomalous customer behavior on an online store: This task relates to machine learning and anomaly detection models, not computer vision.

Suggest automated responses to incoming email: This uses Natural Language Processing (NLP) capabilities, not visual analysis.

Therefore, the correct and Microsoft-verified completion of the statement is:

“Computer vision capabilities can be deployed to integrate a face detection feature into an app.”

The correct answers are B. Import chit-chat content from a predefined data source, C. Manually enter the questions and answers, and E. Generate the questions and answers from an existing webpage.

According to Microsoft Learn and the Azure AI Fundamentals (AI-900) study guide, the Question Answering feature of the Azure AI Language Service (formerly part of QnA Maker) allows developers to create a knowledge base (KB) that enables a chatbot to answer common questions automatically. This knowledge base can be built in three main ways:

Import chit-chat content (B):Azure provides predefined chit-chat datasets that can be imported to make a bot more conversational and natural. This includes small talk such as greetings, acknowledgments, and polite responses (for example, “How are you?” → “I’m doing great, thanks!”). Importing this content enriches the bot’s personality and improves user engagement.

Manually enter questions and answers (C):Developers can manually add pairs of questions and answers directly into the question answering knowledge base. This approach is suitable for custom FAQs or domain-specific content. It gives complete control over how each question is phrased and what answer is returned, ensuring high precision and clarity.

Generate questions and answers from an existing webpage (E):Azure AI Language can automatically extract Q & A pairs from a website’s FAQ or support page. This is done by providing the webpage URL to the service, which scans the page and builds a knowledge base from the detected questions and corresponding answers.

The other options are incorrect:

A (Cortana channel) relates to bot deployment, not knowledge creation.

D (Automated ML) is used for predictive modeling, not for building Q & A datasets.

Thus, the verified correct answers are B, C, and E.

The correct answer is “An embedding.”

In the context of large language models (LLMs) such as GPT-3, GPT-3.5, or GPT-4, an embedding refers to a multi-dimensional numeric vector representation assigned to each word, token, or phrase. According to the Microsoft Azure AI Fundamentals (AI-900) study guide and Microsoft Learn documentation for Azure OpenAI embeddings, embeddings are used to represent textual or semantic meaning in a numerical form that a machine learning model can process mathematically.

Each embedding captures the semantic relationships between words. Words or tokens with similar meanings (for example, “car” and “automobile”) are represented by vectors that are close together in the multi-dimensional space, while unrelated words (like “tree” and “laptop”) are farther apart. This vector representation enables the model to understand context, similarity, and relationships between different pieces of text.

Embeddings are fundamental in tasks such as:

Semantic search: Finding documents or sentences with similar meaning.

Clustering: Grouping related concepts together.

Recommendation systems: Suggesting similar content based on text meaning.

Contextual understanding: Helping generative models produce coherent and context-aware text.

Option review:

Attention: A mechanism used within transformers to focus on relevant parts of input sequences but not a representation of words.

A completion: Refers to the generated text output from a model, not the internal representation.

A transformer: The architecture that powers models like GPT, not the vector representation of tokens.

Therefore, the correct term for a multi-dimensional vector assigned to each word or token in a large language model (LLM) is An embedding, which represents how meaning is numerically encoded and compared within language models.