Welcome to a comprehensive collection of practice questions designed to aid your preparation for the AI-900 exam, your gateway to understanding Azure fundamentals in artificial intelligence.
This repository is thoughtfully divided into two sections. The first section comprises practice questions provided by Microsoft within their training modules. The second section features additional questions curated by me, aligning with various modules covered in the exam syllabus.
To optimise your study experience, it's highly recommended to begin with the official learning modules. Establishing a foundational understanding through these modules will strengthen your grasp of the subject matter, which can then be reinforced and expanded upon with the questions provided here.
Don't feel overwhelmed by the amount of questions! They are organised into modules, ensuring a structured and thorough preparation process.
I have drawn inspiration from this repository
1. You want to create a model to predict sales of ice cream based on historic data that includes daily ice cream sales totals and weather measurements. Which Azure service should you use?
- Azure Machine Learning
- Azure AI Bot Service
- Azure AI Language
Show Answer
Azure Machine Learning
2. You work for a wildlife sanctuary and are considering using AI to identify bird species from images. Which AI service should you use to prototype your idea?
- Azure AI Vision
- Azure AI Search
- Azure OpenAI
Show Answer
Azure AI Vision
3. A predictive app provides audio output for visually impaired users. Which principle of Responsible AI is reflected here?
- Transparency
- Inclusiveness
- Fairness
Show Answer
Inclusiveness
4. You want to create a model to predict the cost of heating an office building based on its size in square feet and the number of employees working there. What kind of machine learning problem is this?
- Regression
- Classification
- Clustering
Show Answer
Regression
- Mean squared error (MSE)
- Precision
- Silhouette
Show Answer
Precision
- To remove data for which no known label values are provided
- To evaluate the aggregate difference between predicted and actual label values
- To calculate the cost of training a neural network rather than a statistical model.Â
Show Answer
To evaluate the aggregate difference between predicted and actual label values
- Automatically deploy new versions of a model as they're trained
- Automatically provision Azure Machine Learning workspaces for new data scientists in an organisation
- Automatically run multiple training jobs using different algorithms and parameters to find the best model
Show Answer
Automatically run multiple training jobs using different algorithms and parameters to find the best model
8. An application requires three separate AI services. To see the cost for each separately, what type of resource(s) should be created?
- A multi-service resource that includes all the AI services
- A single-service resource for each AI service
- It's not possible to see costs for individual AI services
Show Answer
A single-service resource for each AI service
9. After logging into one of the Azure studios, what is one task to complete to begin using the studio?
- Input a key and endpoint into the studio
- Customize the API request.
- Associate a resource with the studio
Show Answer
Associate a resource with the studio
- A bundle of several AI services in one resource
- An AI service to recognize faces
- A single-service resource for Azure AI Search
Show Answer
A bundle of several AI services in one resource
- Timestamps in photograph metadata
- Pixels
- Image file names
Show Answer
Pixels
12. You want to use the Azure AI Vision service to analyze images. You also want to use the Azure AI Language service to analyze text. You want developers to require only one key and endpoint to access all of your services. What kind of resource should you create in your Azure subscription?
- Azure AI Vision
- Azure AI services
- Azure OpenAI service
Show Answer
Azure AI services
13. You want to use the Azure AI Vision service to identify the location of individual items in an image. Which of the following features should you retrieve?
- Objects
- Visual Tags
- Dense Captions
Show Answer
Objects
- A pair of coordinates for each face, indicating the center of the face
- Two pairs of coordinates for each face, indicating the location of the eyes
- A set of coordinates for each face, defining a rectangular bounding box around the face
Show Answer
A set of coordinates for each face, defining a rectangular bounding box around the face
- Glasses
- Extreme angles
- Fast shutter speed
Show Answer
Extreme angles
- Create an Azure Cognitive Search resource, and open Vision Studio
- Create a Face resource, and open Vision Studio
- Create a Face resource, and open Azure OpenAI Studio
Show Answer
Create a Face resource, and open Vision Studio
17. You want to extract text from images and then use Azure AI Language to analyze the text. You want developers to require only one key and endpoint to access all of your services. What kind of resource should you create in your Azure subscription?
- Azure AI Vision
- Azure AI services
- Azure AI Language
Show Answer
Azure AI services
- Results arranged in pages, lines, and words
- Only the bounding box coordinates
- Results arranged by pages that have photographs first, then pages that exclusively have text
Show Answer
Results arranged in pages, lines, and words
19. You want to use Azure AI Language to determine the key talking points in a text document. Which feature of the service should you use?
- Sentiment analysis
- Key phrase extraction
- Entity detection
Show Answer
Key phrase extraction
20. You use Azure AI Language to perform sentiment analysis on a sentence. The confidence scores .04 positive, .36 neutral, and .60 negative are returned. What do these confidence scores indicate about the sentence sentiment?
- The document is positive.
- The document is neutral.
- The document is negative.
Show Answer
The document is negative.
- When the score calculated by the service is outside the range of 0 to 1
- When the predominant language in the text is mixed with other languages
- When the language is ambiguous
Show Answer
When the language is ambiguous
22. Your organization has an existing frequently asked questions (FAQ) document. You need to create a knowledge base that includes the questions and answers from the FAQ with the least possible effort. What should you do?
- Create an empty knowledge base, and then manually copy and paste the FAQ entries into it.
- Import the existing FAQ document into a new knowledge base.
- Import a pre-defined chit-chat data source.
Show Answer
Import the existing FAQ document into a new knowledge base.
23. You want to create a knowledge base for your organization’s bot service. Which Azure AI service is best suited to creating a knowledge base?
- Conversational Language Understanding
- Question Answering
- Optical Character Recognition
Show Answer
Question Answering
24. You need to provision an Azure resource that will be used to author a new conversational language understanding application. What kind of resource should you create?
- Azure AI Speech
- Azure AI Language
- Azure AI services
Show Answer
Azure AI Language
25. You are authoring a conversational language understanding application to support an international clock. You want users to be able to ask for the current time in a specified city, for example "What is the time in London?". What should you do?
- Define a "city" entity and a "GetTime" intent with utterances that indicate the city entity.
- Create an intent for each city, each with an utterance that asks for the time in that city.
- Add the utterance "What time is it in city" to the "None" intent.
Show Answer
Define a "city" entity and a "GetTime" intent with utterances that indicate the city entity.
26. You have published your conversational language understanding application. What information does a client application developer need to get predictions from it?
- The endpoint and key for the application's prediction resource
- The endpoint and key for the application's authoring resource
- The Azure credentials of the user who published the language understanding application
Show Answer
The endpoint and key for the application's prediction resource
27. You plan to build an application that uses Azure AI Speech to transcribe audio recordings of phone calls into text, and then submit the transcribed text to Azure AI Language to extract key phrases. You want to manage access and billing for the application services with a single Azure resource. Which type of Azure resource should you create?
- Speech
- Language
- Azure AI services
Show Answer
Azure AI services
27. You want to use Azure AI Speech service to build an application that reads incoming email message subjects aloud. Which API should you use?
- Speech to text
- Text to speech
- Translator
Show Answer
Text to speech
28. You plan to use Azure AI Document Intelligence's prebuilt receipt model. Which kind of Azure resource should you create?
- Azure AI Vision resource
- Azure AI Document Intelligence or Azure AI services resource.
- Azure AI Language resource.
Show Answer
Azure AI Document Intelligence or Azure AI services resource.
29. You are using the Azure AI Document Intelligence service to analyze receipts. Which field types does the service recognise?
- Merchant retail type.
- Merchant name and address.
- Merchant name and date of incorporation.
Show Answer
Merchant name and address
- Train the model on sample receipts from your organisation.
- Create an Azure AI Document Intelligence resource.
- Nothing - receipt analyzer is available once you create an Azure subscription.
Show Answer
Create an Azure AI Document Intelligence resource.
- CSV
- SQL
- JSON
Show Answer
JSON
- An indexer converts documents into JSON and forwards them to a search engine for indexing.
- An indexer can be used instead of an index if the files are already in the proper format.
- An indexer is only used for AI enrichment and skillset execution.
Show Answer
An indexer converts documents into JSON and forwards them to a search engine for indexing.
33. If you set up a search index without including a skillset, which would you still be able to query?Â
- Sentiment
- Text content
- Image captions
Show Answer
Text content
- Models that only work with one language.
- Models that only work with small amounts of data.
- Models that use deep learning to process and understand natural language on a massive scale.
Show Answer
Models that use deep learning to process and understand natural language on a massive scale.
- Monitoring the temperature in a manufacturing facility.
- Creating a draft for an email.
- Collecting real time data and storing it in a database.
Show Answer
Creating a draft for an email.
- To represent semantic meaning of text tokens.
- To create tokens that include multiple representations of a word in different languages.
- To correct misspellings in the training data.
Show Answer
To represent semantic meaning of text tokens.
- Copilots only impact applications used in professional settings.
- Copilots can help with first drafts, information synthesis, strategic planning, and much more.
- Copilots can only be used for certain natural language tasks like summarizing text.
Show Answer
Copilots can help with first drafts, information synthesis, strategic planning, and much more.
39. Which assumption of the multiple linear regression model should be satisfied to avoid misleading predictions?
- Features are dependent on each other.
- Features are independent of each other.
- Labels are dependent on each other.
- Labels are independent of each other.
Show Answer
Features are independent of each other.
- Azure OpenAI is Microsoft's version of ChatGPT, a chatbot that uses generative AI models.
- ChatGPT and OpenAI are chatbots that generate natural language, code, and images. Azure OpenAI provides access to these two chatbots.
- OpenAI is a research company that developed ChatGPT, a chatbot that uses generative AI models. Azure OpenAI provides access to many of OpenAI's AI models
Show Answer
OpenAI is a research company that developed ChatGPT, a chatbot that uses generative AI models. Azure OpenAI provides access to many of OpenAI's AI models
41. You would like to summarise a paragraph of text. Which generative AI model family would you use to solve for this workload?Â
- GPT
- Codex
- Dall-E.
Show Answer
GPT
- Provides Transparency Notes that share how technology is built and asks users to consider its implications.
- Logs users out of Azure OpenAI Studio after a period of inactivity to ensure it's only used by one user.
- Allows users to build any application, regardless of harmful effects, to ensure fairness.
Show Answer
Provides Transparency Notes that share how technology is built and asks users to consider its implications.
43. Why should you consider creating an AI Impact Assessment when designing a generative AI solution?
- To make a legal case that indemnifies you from responsibility for harms caused by the solution
- To document the purpose, expected use, and potential harms for the solution
- To evaluate the cost of cloud services required to implement your solution
Show Answer
To document the purpose, expected use, and potential harms for the solution
44. What capability of Azure OpenAI Service helps mitigate harmful content generation at the Safety System level?
- DALL-E model support
- Fine-tuning
- Content filters
Show Answer
Content filters
- To enable you to gather feedback and identify issues before releasing the solution more broadly
- To eliminate the need to identify, measure, and mitigate potential harms
- To enable you to charge more for the solution
Show Answer
To enable you to gather feedback and identify issues before releasing the solution more broadly
- Software that imitates human behaviors and capabilities
- Advanced mathematics
- Machine learning, computer vision, natural language processing
- Programming language
Show Answer
Software that imitates human behaviors and capabilities
- Data science
- Advanced mathematics
- Machine learning, computer vision, natural language processing
- Robotics
Show Answer
Machine learning, computer vision, natural language processing
- By watching humans
- Through trial and error
- Through data
- By reading books
Show Answer
Through data
- In agriculture
- In robotics
- In healthcare
- In Microsoft Azure
Show Answer
In Microsoft Azure
- Natural language processing
- Robotics
- Computer vision
- Machine learning
Show Answer
Computer vision
- A mechanism for visual perception
- A process of naming images
- A method for classifying images based on their contents
- An array of pixel values
Show Answer
A method for classifying images based on their contents
- Reliability and safety
- Privacy and security
- Fairness
- Inclusiveness
Show Answer
Fairness
- Reliability and safety
- Transparency
- Privacy and security
- Accountability
Show Answer
Privacy and security
- Computer vision
- Machine learning
- Natural language processing
- Generative AI
Show Answer
Natural language processing
- Azure Vision Studio
- Azure AI Language
- Azure OpenAI Service
- Azure AI Vision
Show Answer
Azure OpenAI Service
- To write code for predictive models
- To use data to create a predictive model
- To study historical data only
- To predict past observations accurately
Show Answer
To use data to create a predictive model
2. What is the process called when a machine learning model calculates an output value based on one or more input values?
- Inference
- Training
- Collaboration
- Modeling
Show Answer
Inference
- Features represent observed attributes, and labels represent predictions
- Features represent predictions, and labels represent observed attributes
- Features represent historical data, and labels represent future data
- Features represent future data, and labels represent historical data
Show Answer
Features represent observed attributes, and labels represent predictions
- To classify observations into discrete groups
- To predict a numeric value
- To predict a categorization or class
- To determine relationships between features
Show Answer
To predict a numeric value
- To determine relationships between features and labels
- To predict a categorization or class
- To group observations based on similarities
- To identify patterns in historical data
Show Answer
To group observations based on similarities
- A form of unsupervised learning
- A form of supervised learning where the label represents a categorization
- A form of supervised learning where the label predicted by the model is a numeric value
- A form of machine learning used for clustering
Show Answer
A form of supervised learning where the label predicted by the model is a numeric value
- Predicting whether a bank customer will default on a loan
- Predicting the genre of a movie
- Predicting the species of a penguin
- Predicting the fuel efficiency of a car
Show Answer
Predicting the fuel efficiency of a car
- A form of unsupervised learning
- A form of supervised learning where the label represents a categorization
- A form of supervised learning where the label predicted by the model is a numeric value
- A form of machine learning used for clustering
Show Answer
A form of supervised learning where the label represents a categorization
- Predicting the species of a penguin
- Predicting the selling price of a property
- Predicting whether a patient is at risk for diabetes
- Predicting the fuel efficiency of a car
Show Answer
Predicting whether a patient is at risk for diabetes
- Accuracy
- Recall
- Mean Absolute Error (MAE)
- Precision
Show Answer
Mean Absolute Error (MAE)
- Mean Squared Error (MSE)
- F1-score
- Root Mean Squared Error (RMSE)
- Coefficient of determination (R2)
Show Answer
F1-score
- A form of unsupervised learning
- A form of supervised learning where the label represents a categorization
- A form of supervised learning where the label predicted by the model is a numeric value
- A form of machine learning used for clustering
Show Answer
A form of supervised learning where the label represents a categorization
- Predicting the selling price of a property
- Predicting whether a bank customer will default on a loan
- Predicting the genre of a movie
- Predicting the fuel efficiency of a car
Show Answer
Predicting the genre of a movie
- A form of supervised learning where the label represents a categorization
- A form of unsupervised learning
- A form of machine learning used for regression
- A form of machine learning used for classification
Show Answer
A form of unsupervised learning
- Predicting the species of a penguin
- Grouping flowers based on the number of leaves and petals
- Predicting the fuel efficiency of a car
- Predicting whether a patient is at risk for diabetes
Show Answer
Grouping flowers based on the number of leaves and petals
- To maximize accuracy
- To minimize recall
- To minimize the mean absolute error (MAE)
- To group data points into clusters based on similarities
Show Answer
To group data points into clusters based on similarities
- Accuracy
- Recall
- Silhouette
- Precision
Show Answer
Silhouette
- A form of unsupervised learning
- A form of supervised learning
- A form of machine learning used for regression
- An advanced form of machine learning inspired by the human brain
Show Answer
An advanced form of machine learning inspired by the human brain
- A biological system for transmitting information
- An algorithm for clustering data points
- A function that maps inputs to outputs based on learned weights
- A type of activation function
Show Answer
A function that maps inputs to outputs based on learned weights
- A programming language for machine learning
- A cloud service for managing machine learning projects
- An open-source machine learning library
- A deep learning framework
Show Answer
A cloud service for managing machine learning projects
- To import and explore data
- To provision compute resources
- To run code in notebooks
- To manage machine learning resources and jobs
Show Answer
To manage machine learning resources and jobs
- They require specialist AI knowledge to implement
- They are only accessible to large technology companies
- They unlock automation for workloads in language, vision, intelligent search, content generation, and more
- They are primarily designed for robotics applications
Show Answer
They unlock automation for workloads in language, vision, intelligent search, content generation, and more
- By requiring extensive coding skills
- By using pre-trained machine learning models and deploying them as resources
- By providing specialized APIs for each service
- By limiting access to only large organizations
Show Answer
By using pre-trained machine learning models and deploying them as resources
- Basic and advanced
- Public and private
- Multi-service and single-service
- Internal and external
Show Answer
Multi-service and single-service
- Through REST APIs, client libraries, or integrated tools
- Through complex programming languages only
- Only by using Logic Apps and Power Automate
- Through manual configuration in the Azure portal only
Show Answer
Through REST APIs, client libraries, or integrated tools
- To provide access to free resources
- To verify the user's identity and authorization to use the service
- To restrict access to specific geographical locations
- To ensure compatibility with legacy systems
Show Answer
To verify the user's identity and authorization to use the service
- A representation of biological eyes
- A visual perception mechanism
- An array of numeric pixel values
Show Answer
An array of numeric pixel values
- RGB color hues
- Grayscale image
- Object detection
Show Answer
Grayscale image
- To remove data from images
- To manipulate the size of images
- To modify pixel values and create visual effects
Show Answer
To modify pixel values and create visual effects
4. Which machine learning model architecture is commonly used in computer vision for image classification?
- Recurrent Neural Networks (RNNs)
- Convolutional Neural Networks (CNNs)
- Transformer Networks
Show Answer
Convolutional Neural Networks (CNNs)
- Randomly
- Based on image labels
- Manually
Show Answer
Based on image labels
- Encoding language tokens as vector-based embeddings
- Removing noise from textual data
- Generating image captions
Show Answer
Encoding language tokens as vector-based embeddings
- To provide prebuilt and customizable computer vision models
- To train language models
- To perform image editing tasks
Show Answer
To provide prebuilt and customizable computer vision models
- Generating captions and descriptions of images
- Only optical character recognition
- Image compression
Show Answer
Generating captions and descriptions of images
- To analyze text data
- To locate and analyze human faces in images or video content
- To identify animals in photographs
- To generate captions for images
Show Answer
To locate and analyze human faces in images or video content
- Identifying animals in photographs
- Automatically tagging known friends in photographs on social media
- Detecting objects in images
- Analyzing text data
Show Answer
Automatically tagging known friends in photographs on social media
- To provide basic face detection and analysis
- To identify animals in videos
- To analyze text data
- To generate captions for images
Show Answer
To provide basic face detection and analysis
- Species of animals
- Facial expressions
- Accessories, blur, exposure, glasses, head pose, mask, noise, and occlusion
- Text content
Show Answer
Accessories, blur, exposure, glasses, head pose, mask, noise, and occlusion
- Submit an intake form
- Subscribe to a premium plan
- Upgrade to a higher-tier Azure subscription
- Enroll in a certification program
Show Answer
Submit an intake form
- Optical Character Rendering
- Object Character Recognition
- Optical Character Recognition
- Object Code Recognition
Show Answer
Optical Character Recognition
- Digitizing medical records
- Scanning checks for bank deposits
- Analyzing satellite images
- Note-taking
Show Answer
Analyzing satellite images
- Extract machine-readable text from images, PDFs, and TIFF files
- Analyze sentiment in text data
- Detect objects in images
- Generate synthetic images
Show Answer
Extract machine-readable text from images, PDFs, and TIFF files
- Vision Studio
- REST API
- GraphQL
- Software Development Kits (SDKs)
Show Answer
GraphQL
- Writing code to interact with Azure AI Vision services
- Providing a graphical interface to access Azure AI Vision APIs
- Analyzing data from OCR results
- Training machine learning models
Show Answer
Providing a graphical interface to access Azure AI Vision APIs
6. Which resource type in Azure subscription allows access to multiple Azure AI services with one key and endpoint?
- Azure AI Vision
- Azure AI services
- Azure AI Language
- Azure AI Compute
Show Answer
Azure AI services
- Pages -> Text blocks -> Words
- Pages -> Lines -> Words
- Pages -> Paragraphs -> Sentences
- Pages -> Regions -> Text blocks
Show Answer
Pages -> Lines -> Words
- Reducing the need for manual data entry
- Increasing hardware efficiency
- Improving network security
- Accelerating graphic rendering
Show Answer
Reducing the need for manual data entry
- Marking areas of text in an image
- Highlighting areas of interest in an image
- Identifying specific objects in an image
- Measuring the aspect ratio of an image
Show Answer
Marking areas of text in an image
- To interpret the subject of text in a way similar to humans
- To encrypt text for secure transmission
- To compress text files for storage efficiency
- To convert text to speech
Show Answer
To interpret the subject of text in a way similar to humans
- Generating random text strings
- Analyzing satellite images
- Creating a social media sentiment analyzer
- Converting images to PDFs
Show Answer
Creating a social media sentiment analyzer
- Breaking down a text into distinct words or phrases
- Encrypting text for secure transmission
- Converting text to speech
- Analyzing the frequency of words in a text
Show Answer
Breaking down a text into distinct words or phrases
4. Which of the following techniques is used to determine the most commonly used words in a document?
- Tokenization
- Stemming
- Frequency analysis
- Entity recognition
Show Answer
Frequency analysis
- To break down text into distinct words or phrases
- To detect the language of a text
- To identify people, places, events, and more in text
- To summarize the main points of a document
Show Answer
To identify people, places, events, and more in text
6. Which Azure resource type should you choose if you want to manage access and billing for Azure AI Language services separately from other services?
- Azure AI Vision
- Language resource
- Azure AI services resource
- Azure Compute
Show Answer
Language resource
- To identify people, places, events, and more in text
- To break down text into distinct words or phrases
- To detect positive, negative, or neutral sentiments in text
- To summarize the main points of a document
Show Answer
To detect positive, negative, or neutral sentiments in text
- The sentiment of a text
- The language of a text
- The main points of a document
- The entities in a text
Show Answer
The language of a text
- To encrypt text for secure transmission
- To detect the language of a text
- To identify the main points or concepts in a document
- To summarize the main points of a document
Show Answer
To identify the main points or concepts in a document
- Image recognition
- Sentiment analysis
- Video processing
- Audio transcription
Show Answer
Sentiment analysis
- A solution that enables dialogue between an AI agent and a human
- A technique to encrypt communication between computers
- A method for compressing large text files
- An approach to convert speech into text
Show Answer
A solution that enables dialogue between an AI agent and a human
- To generate random questions
- To provide automated responses to customer queries
- To encrypt communication channels
- To compress large text files
Show Answer
To provide automated responses to customer queries
- Azure AI Computer Vision and Azure Speech Services
- Azure AI Language and Azure AI Bot Service
- Azure Compute and Azure IoT Hub
- Azure Machine Learning and Azure Cognitive Search
Show Answer
Azure AI Language and Azure AI Bot Service
- Define questions and answers
- Provision a Language resource in Azure
- Test the project
- Build a bot with Azure AI Bot Service
Show Answer
Provision a Language resource in Azure
- To encrypt communication between the bot and users
- To establish a secure connection with the Azure AI Bot Service
- To enable users to interact with the bot through various communication media
- To compress large text files for efficient transmission
Show Answer
To enable users to interact with the bot through various communication media
- An example of something a user might say that the application must interpret
- A method for training language models
- A technique to encrypt communication channels
- A type of Azure resource
Show Answer
An example of something a user might say that the application must interpret
- A pre-defined intent for common scenarios
- An item to which an utterance refers
- A method for connecting client applications to prediction resources
- A natural language expression that a user might say
Show Answer
An item to which an utterance refers
- A specific instance of a general device entity
- The purpose or goal expressed in a user's utterance
- A resource for training language models
- A type of pre-built domain
Show Answer
The purpose or goal expressed in a user's utterance
- To handle utterances that do not match any other intent
- To define pre-built intents for common scenarios
- To train language models
- To create entities for utterances
Show Answer
To handle utterances that do not match any other intent
- To generate random sample utterances
- To define entities, intents, and utterances for training
- To connect client applications to prediction resources
- To publish language models for client consumption
Show Answer
To define entities, intents, and utterances for training
- Speech recognition and transcription
- Speech synthesis and natural language processing
- Speech recognition and synthesis
- Speech analysis and sentiment detection
Show Answer
Speech recognition and synthesis
- To convert text to speech
- To detect and interpret spoken input
- To analyze sentiment in spoken conversations
- To generate captions for audio files
Show Answer
To detect and interpret spoken input
- Phonetic and grammatical models
- Acoustic and language models
- Syntax and semantic models
- Feature extraction and classification models
Show Answer
Acoustic and language models
- To detect spoken patterns in audio files
- To generate phonetic transcriptions
- To convert text to spoken output
- To analyze prosodic features in speech
Show Answer
To convert text to spoken output
- Azure Compute
- Azure Storage
- Azure AI services
- Azure Networking
Show Answer
Azure AI services
- To create documents
- To process text and extract information from documents
- To share documents
- To archive documents
Show Answer
To process text and extract information from documents
- By manually entering data into a database
- By converting scanned images of receipts into text using OCR
- By outsourcing the task to human workers
- By printing receipts on heat-sensitive paper
Show Answer
By converting scanned images of receipts into text using OCR
- Simple and complex models
- Prebuilt and custom models
- Online and offline models
- Static and dynamic models
Show Answer
Prebuilt and custom models
4. What kind of information can prebuilt models in Azure AI Document Intelligence extract from invoices?
- Only customer names
- Only transaction details
- Only sales details
- Customer and vendor details, sales and transaction details, and more
Show Answer
Customer and vendor details, sales and transaction details, and more
5. What is the main advantage of using the prebuilt receipt model in Azure AI Document Intelligence?
- It can process images only in JPEG format
- It can process images of any format and size
- It can process only one receipt type
- It can recognize data on various receipt types with different languages and formats
Show Answer
It can recognize data on various receipt types with different languages and formats
- To manually read through documents
- To automatically extract information from unstructured data
- To organize documents in a file system
- To create documents
Show Answer
To automatically extract information from unstructured data
- Support for only one language
- Integration with Office 365 only
- Support for image and text analysis
- Restricted to cloud assets only
Show Answer
Support for image and text analysis
- By restricting data sources to Azure databases only
- By providing built-in connectors for data ingestion
- By allowing only JSON format data
- By requiring manual data entry
Show Answer
By providing built-in connectors for data ingestion
- To define search queries
- To store search results
- To automate data ingestion
- To apply a sequence of AI skills to enrich data
Show Answer
To apply a sequence of AI skills to enrich data
- As rows in a database
- As JSON documents
- As tables in a file system
- As columns in a spreadsheet
Show Answer
As JSON documents
6. What are the two approaches provided by Azure AI Search to create and load JSON documents into an index?
- Push method and Pull method
- Export method and Import method
- REST API method and SDK method
- Source method and Target method
Show Answer
Push method and Pull method
7. Which method does Azure AI Search use to pull data from external Azure data sources and populate a search index?
- Push method
- REST API method
- Pull method
- Import method
Show Answer
Pull method
- By checking the Azure AI Search dashboard
- By monitoring the associated indexer
- By using the Search explorer tool
- All of the above
Show Answer
All of the above
- To persist connection information to source data
- To store enriched content generated from AI skillsets
- To create a physical data structure for full text search
- To manage components of the Azure AI Search resource
Show Answer
To store enriched content generated from AI skillsets
- Simple query syntax
- Full Lucene query syntax
- Advanced query syntax
- Complex query syntax
Show Answer
Simple query syntax
- AI that imitates human behavior using predefined instructions
- AI that interacts with the environment without explicit directions
- AI that generates original content
- AI that analyzes existing data and makes predictions
Show Answer
AI that generates original content
- Weather forecasting
- Stock market analysis
- ChatGPT, a chatbot
- Medical diagnosis
Show Answer
ChatGPT, a chatbot
- Text only
- Text, images, and code
- Images only
- Code only
Show Answer
Text, images, and code
- To create natural language responses
- To decompose training text into tokens
- To generate embeddings for tokens
- To calculate attention scores
Show Answer
To decompose training text into tokens
- To predict the next token in a sequence
- To evaluate the semantic relationships between tokens
- To decompose training text into tokens
- To generate embeddings for tokens
Show Answer
To evaluate the semantic relationships between tokens
- Microsoft's cloud solution for deploying, customizing, and hosting large language models
- A platform for weather forecasting
- An API for stock market analysis
- A medical diagnosis tool
Show Answer
Microsoft's cloud solution for deploying, customizing, and hosting large language models
- Deploy large language models and provide few-shot examples
- Build and deploy weather forecasting models
- Train medical diagnosis models
- Analyze stock market trends
Show Answer
Deploy large language models and provide few-shot examples
- Large language models for generating code
- AI assistants that help users with common tasks
- Models specifically designed for medical diagnosis
- Tools for weather forecasting
Show Answer
AI assistants that help users with common tasks
- The process of designing user interfaces
- The process of improving the quality of responses from generative AI by refining prompts
- The process of optimizing search engine algorithms
- The process of training machine learning models
Show Answer
The process of improving the quality of responses from generative AI by refining prompts
- By using complex prompts
- By providing examples and grounding data in prompts
- By increasing the size of the model
- By reducing the amount of training data
Show Answer
By providing examples and grounding data in prompts
- To utilize Azure's infrastructure, including security, compliance, and regional availability, to help users build enterprise-grade applications.
- To deploy OpenAI AI model capabilities across Microsoft products, including and beyond Azure AI products.
- To use Azure to power all of OpenAI's workloads.
- All of the above
Show Answer
All of the above
- Understanding and creating natural language
- Generating images
- Translating code from one programming language into another
- None of the above
Show Answer
Understanding and creating natural language
- By registering for limited access
- By downloading the Azure OpenAI Studio app
- By purchasing a subscription from the Microsoft Store
- All of the above
Show Answer
By registering for limited access
- By providing code suggestions and completions
- By automatically debugging code
- By generating entire software applications
- None of the above
Show Answer
By providing code suggestions and completions
- They ensure that only Microsoft employees can access the service
- They protect users' AI models and data
- They prioritize speed over security
- They limit the functionality of the AI models
Show Answer
They protect users' AI models and data
- By allowing users to build any application without restrictions
- By providing tools to detect and mitigate harmful use cases
- By promoting biased decision-making
- None of the above
Show Answer
By providing tools to detect and mitigate harmful use cases
- To explain how to use the Azure OpenAI Studio app
- To outline the terms of service for Azure OpenAI
- To share information about how the technology is built and its implications
- None of the above
Show Answer
To share information about how the technology is built and its implications
1. According to Microsoft's guidelines, what is the first stage in developing a responsible generative AI plan?
- Identify potential harms
- Measure potential harms
- Mitigate potential harms
Show Answer
Identify potential harms
2. Which layer of a generative AI solution focuses on the construction of prompts submitted to the model?
- The model layer
- The safety system layer
- The metaprompt and grounding layer
Show Answer
The metaprompt and grounding layer
3. What is the purpose of the "red team" testing mentioned in the responsible generative AI process?
- To verify the presence of potential harms in the solution
- To prioritize the identified harms
- To test the robustness of the model layer
Show Answer
To verify the presence of potential harms in the solution
- Performing only manual testing to ensure accuracy
- Using pre-defined criteria to evaluate the output
- Applying harm categorization after the testing phase
Show Answer
Using pre-defined criteria to evaluate the output
- Descriptions of the user interface design only
- Information about the generative AI model used
- Recommendations for system deployment
Show Answer
Information about the generative AI model used
- To fine-tune the generative AI model
- To apply platform-level configurations and capabilities to mitigate harm
- To design the user interface of the application
Show Answer
To apply platform-level configurations and capabilities to mitigate harm
7. What is the significance of devising a phased delivery plan for releasing a generative AI solution?
- It allows for rapid deployment without user feedback
- It helps gather feedback and identify problems before a wider release
- It ensures complete secrecy of the solution until full release
Show Answer
It helps gather feedback and identify problems before a wider release