11 free Microsoft AB-731 practice questions from the CertStash pack of 100, each with the correct answer and a full explanation. The rule is the same for every exam: ten free on the product page, 11 free here, all 100 in the pack, and each step includes the one before it.
Get all 100 questions (US$39) · Download the first 10 as a PDF
Question 1
Which statement accurately describes the difference between a pretrained generative AI model and a fine-tuned generative AI model?
Show answer and explanation
Correct answer: C. A pretrained model is trained on broad datasets, while a fine-tuned model is adapted to perform well on a narrower, domain-specific dataset.
A pretrained generative AI model is trained on large, broad datasets to learn general patterns and capabilities across diverse domains. A fine-tuned model takes a pretrained model and adapts it through additional training on a smaller, domain-specific dataset to optimize performance for a particular task or industry. This approach leverages the general knowledge from pretraining while specializing the model for narrower use cases.
Why the other options are wrong
- A. Both pretrained and fine-tuned models can require labeled data depending on the training approach used.
- B. Pretrained models typically use more parameters and take longer to train initially; fine-tuning is faster because it builds on existing weights rather than training from scratch.
- D. This reverses the actual definitions, pretrained models are general-purpose, while fine-tuned models are optimized for specific tasks.
Question 2
Your company receives thousands of scanned invoices each month. You need to recommend an AI solution that can automatically extract key details, such as invoice numbers, vendor names, and total amounts. What is the best solution to recommend? More than one answer choice may achieve the goal. Select the BEST answer.
Show answer and explanation
Correct answer: A. Azure Document Intelligence in Foundry Tools
Azure Document Intelligence is purpose-built for extracting structured data from documents like invoices, receipts, and forms. It can automatically identify and extract key fields such as invoice numbers, vendor names, dates, and amounts using specialized models trained for document understanding. This is the most appropriate and efficient solution for automated invoice processing at scale.
Why the other options are wrong
- B. Azure Vision is designed for image analysis and object detection but lacks the specialized document structure understanding needed for accurate field extraction from invoices.
- C. Azure AI Search is an indexing and retrieval service optimized for full-text search and information discovery, not for structured data extraction from document images.
- D. Azure Machine Learning is a general-purpose platform for building custom models; it would require significant effort to build a custom solution when a purpose-built service exists.
Question 3
Your company discovers that several employees use personal ChatGPT accounts to assist with work tasks. You are concerned about proprietary data being shared externally. You need to evaluate the business value of rolling out Microsoft 365 Copilot. Which capability is a key benefit of using Copilot instead of a personal ChatGPT account?
Show answer and explanation
Correct answer: D. accessing internal data in accordance with existing Microsoft 365 policies
The primary security and governance advantage of Microsoft 365 Copilot over personal ChatGPT accounts is that it respects existing Microsoft 365 permission policies and data access controls. Copilot can safely access internal organizational data, documents, emails, files, while enforcing the same security boundaries and compliance requirements already in place, preventing data from being shared with external services or exposed inappropriately.
Why the other options are wrong
- A. While Copilot can analyze and produce reports, this capability is not unique to Copilot compared to ChatGPT and does not address the security concern of proprietary data leakage.
- B. Both ChatGPT and Copilot can generate ideas and solve issues; this capability does not distinguish between them in terms of data security.
- C. Both ChatGPT and Copilot can draft documents and marketing materials; this is not the differentiating factor that addresses the risk of proprietary data being shared externally.
Question 4
Your company plans to build a generative AI solution based on internal data. You recommend using Microsoft Foundry as a starting point to develop and manage the solution. What is a key benefit of using Microsoft Foundry for this project?
Show answer and explanation
Correct answer: A. Provides a scalable platform for developing and deploying generative AI solutions.
Microsoft Foundry provides a scalable platform for developing and deploying generative AI solutions, enabling organizations to build end-to-end AI applications with integrated tools and services. This addresses the core need stated in the question of building and managing a generative AI solution based on internal data.
Why the other options are wrong
- B. Foundry still requires selection and configuration of underlying AI models; it does not eliminate this step.
- C. While Foundry can support business users to some degree, its primary value is as a comprehensive development platform for AI solutions, not specifically enabling business users without technical background.
- D. Foundry is not primarily characterized as a low-code platform; it provides a full development environment that supports multiple approaches.
Question 5
Which business requirement most closely relates to grounding a generative AI model?
Show answer and explanation
Correct answer: D. ensuring that verified company data sources are used for response generation
Grounding a generative AI model means anchoring its responses to specific, verified data sources within the organization. This ensures that the model generates responses based on trusted company data rather than relying solely on its training data, which directly relates to ensuring verified company data sources are used for response generation.
Why the other options are wrong
- A. Supporting multiple languages is a localization feature unrelated to grounding.
- B. Measuring user interaction metrics relates to usage analytics, not to grounding model responses.
- C. Natural language query capability is a feature of generative AI models but not specifically related to grounding.
Question 6
You need to create a custom Azure Machine Learning model. The data used to train the model is consistent and uniform. What should you do first?
Show answer and explanation
Correct answer: A. Prepare the training data.
Preparing the training data is the foundational first step in creating any Azure Machine Learning model. Even with consistent and uniform data, data preparation ensures proper formatting, validation, and organization before any model training can begin. This step must precede training, evaluation, hyperparameter tuning, and deployment.
Why the other options are wrong
- B. Evaluation occurs after model training, not before.
- C. Training cannot occur without prepared data to train on.
- D. Hyperparameter tuning occurs during or after initial model training.
- E. Deployment occurs only after the model has been trained and evaluated.
Question 7
Your company uses a non-reasoning generative AI model to create textual content. You discover that the model’s responses are inconsistent and do NOT meet expectations. You need to improve the prompts. What should you do? More than one answer choice may achieve the goal. Select the BEST answer.
Show answer and explanation
Correct answer: B. Add the context, sources, and expectations to the prompts.
Adding context, sources, and expectations to prompts provides the generative AI model with the necessary information to produce more consistent and accurate responses. This comprehensive approach addresses the root cause of inconsistent outputs by giving the model clear boundaries and reference points for generating content.
Why the other options are wrong
- A. While examples can help, they are less effective than providing context and sources, which directly ground the model's responses.
- C. Technical jargon can confuse generative AI models rather than enhance comprehension; clarity is preferred over technical density.
- D. A single concise requirement is insufficient for addressing inconsistent responses; more comprehensive guidance is needed.
Question 8
Your company plans to use generative AI to help project managers and engineers work with construction blueprints stored as PDF files. You need to recommend a generative AI solution that meets the following business requirements: Processes both images and text Summarizes the design of a building Answers user questions about a building’s design Extracts information from blueprints, such as the location of electrical, heating, and plumbing systems What should you recommend?
Show answer and explanation
Correct answer: A. a multi-modal solution
A multi-modal solution processes both images and text simultaneously, making it ideal for construction blueprints stored as PDFs. This capability enables the system to analyze the visual blueprint components while extracting and understanding textual information, and then perform summarization and answer questions, all requirements specified in the scenario.
Why the other options are wrong
- B. OCR solutions only extract text from images; they do not process both modalities or provide summarization and question-answering capabilities.
- C. Text completion solutions generate text continuations and do not process images or extract information from visual blueprints.
- D. Document summarization alone does not address the image processing requirement or the ability to extract specific infrastructure information from visual elements.
Question 9
In which scenario is Azure Machine Learning most likely to deliver strategic value for an organization?
Show answer and explanation
Correct answer: A. Using historical sales data to forecast demand across product categories.
Using historical sales data to forecast demand across product categories represents a predictive analytics use case where Azure Machine Learning delivers significant strategic value through improved inventory management, revenue optimization, and resource planning. This is a core machine learning application that drives business decisions.
Why the other options are wrong
- B. Digitizing paper processes primarily improves operational efficiency through automation, not through advanced machine learning capabilities.
- C. Entering customer feedback into a spreadsheet for sentiment understanding is a manual data entry task that does not leverage Azure Machine Learning's analytical power.
- D. Sending location-based personalized emails is a marketing automation task that requires segmentation logic rather than sophisticated machine learning models.
Question 10
Your company uses a generative AI solution. You need to improve the quality of responses by using grounding. Which statement accurately describes how grounding improves accuracy and relevancy?
Show answer and explanation
Correct answer: C. anchors the responses in specific data sources
Grounding anchors generative AI responses in specific, verified data sources, which directly improves accuracy and relevancy by ensuring the model references factual information rather than relying solely on patterns learned during training. This technique reduces hallucinations and increases the trustworthiness of generated content.
Why the other options are wrong
- A. Referencing diverse perspectives relates to input diversity, not to the grounding mechanism that anchors responses in specific data sources.
- B. Explaining how and why models generate content addresses interpretability, not the accuracy and relevancy improvements provided by grounding.
- D. Specifying model strengths and weaknesses is about model documentation, not about the grounding technique that improves response quality.
What people say about the full AB-731 pack 5.0 from 1
You are the best site I have found for AI Transformation Leader dump
Question 11
You plan to meet with a group of stakeholders to discuss how generative AI can benefit your company. You need to provide the stakeholders with a relevant description of generative AI during the meeting. Which description should you use?
Show answer and explanation
Correct answer: A. Generative AI is designed to generate responses based on a user’s natural language prompts.
Generative AI is fundamentally designed to generate responses based on user prompts expressed in natural language. This is the defining characteristic of generative AI systems like large language models, which distinguish them from other AI approaches. This description provides stakeholders with an accurate, accessible understanding of what generative AI does.
Why the other options are wrong
- B. Translation is a specialized natural language processing task, not the core purpose of generative AI.
- C. Predicting future trends based on historical data is a predictive analytics function, not a generative AI capability.
- D. Product recommendations based on user behavior are performed by recommender systems, not generative AI models.
That was 11 of 100.
The full Microsoft AB-731 pack has all 100 questions, each with the answer, the explanation and why the other options are wrong, plus a questions-only copy for timed runs. US$39, paid once, with free monthly updates and a pass-or-your-money-back guarantee.
Not ready yet? Get the first 11 free as one PDF by email. Get the Half Pack →