An AI practitioner trained a custom model on Amazon Bedrock by using a training dataset that contains confidential data. The AI practitioner wants to ensure that the custom model does not generate inference responses based on confidential data.
How should the AI practitioner prevent responses based on confidential data?
- A . Delete the custom model. Remove the confidential data from the training dataset. Retrain the custom model.
- B . Mask the confidential data in the inference responses by using dynamic data masking.
- C . Encrypt the confidential data in the inference responses by using Amazon SageMaker.
- D . Encrypt the confidential data in the custom model by using AWS Key Management Service (AWS KMS).
A
Explanation:
A: Delete the custom model. Remove the confidential data from the training dataset. Retrain the custom model. Explanation: If the training dataset contains confidential data, the model may inadvertently learn and generate responses based on that data. The only way to ensure that the model does not generate responses based on the confidential data is to: Remove the confidential data from the training dataset. Retrain the custom model using the updated dataset. This process ensures that the model is not influenced by the sensitive information.
Which feature of Amazon OpenSearch Service gives companies the ability to build vector database applications?
- A . Integration with Amazon S3 for object storage
- B . Support for geospatial indexing and queries
- C . Scalable index management and nearest neighbor search capability
- D . Ability to perform real-time analysis on streaming data
C
Explanation:
The Amazon OpenSearch Service supports building vector database applications by enabling nearest neighbor search capability. This feature allows the service to efficiently perform similarity searches, which is crucial for applications that rely on vector embeddings (e.g., recommendation systems, image or text similarity searches). Combined with scalable index management, this makes OpenSearch an excellent choice for vector database applications.
A company wants to display the total sales for its top-selling products across various retail locations in the past 12 months.
Which AWS solution should the company use to automate the generation of graphs?
- A . Amazon Q in Amazon EC2
- B . Amazon Q Developer
- C . Amazon Q in Amazon QuickSight
- D . Amazon Q in AWS Chatbot
C
Explanation:
Amazon Q is a feature within Amazon QuickSight that allows users to ask questions about their data in natural language and receive visualizations as responses. This functionality is particularly useful for generating graphs and visualizations based on specific queries regarding sales data.
A company wants to build an interactive application for children that generates new stories based on classic stories. The company wants to use Amazon Bedrock and needs to ensure that the results and topics are appropriate for children.
Which AWS service or feature will meet these requirements?
- A . Amazon Rekognition
- B . Amazon Bedrock playgrounds
- C . Guardrails for Amazon Bedrock
- D . Agents for Amazon Bedrock
C
Explanation:
C – Guardrails for Amazon Bedrock provides the necessary tools to ensure that the interactive story-generating application remains safe, appropriate, and engaging for children, making it the best choice for this scenario.
A company has developed an ML model for image classification. The company wants to deploy the model to production so that a web application can use the model.
The company needs to implement a solution to host the model and serve predictions without managing any of the underlying infrastructure.
Which solution will meet these requirements?
- A . Use Amazon SageMaker Serverless Inference to deploy the model.
- B . Use Amazon CloudFront to deploy the model.
- C . Use Amazon API Gateway to host the model and serve predictions.
- D . Use AWS Batch to host the model and serve predictions.
A
Explanation:
Amazon SageMaker Serverless Inference is a fully managed solution for deploying machine learning models without managing the underlying infrastructure. It automatically provisions compute capacity, scales based on request traffic, and serves predictions efficiently. This makes it an ideal choice for hosting a model and serving predictions for a web application with minimal management overhead. Why not the other options? B: Use Amazon CloudFront to deploy the model: Amazon CloudFront is a content delivery network (CDN) C: Use Amazon API Gateway to host the model and serve predictions: Amazon API Gateway is used to create APIs for accessing services. D: Use AWS Batch to host the model and serve predictions: AWS Batch is designed for batch processing and job scheduling, not for real-time inference or hosting ML models for web applications
A company has petabytes of unlabeled customer data to use for an advertisement campaign. The
company wants to classify its customers into tiers to advertise and promote the company’s products.
Which methodology should the company use to meet these requirements?
- A . Supervised learning
- B . Unsupervised learning
- C . Reinforcement learning
- D . Reinforcement learning from human feedback (RLHF)
B
Explanation: Unsupervised learning is used when working with unlabeled data, such as the customer data described in this scenario. This methodology allows the company to identify patterns and group similar customers into clusters or tiers without the need for predefined labels. Techniques like clustering (e.g., K-Means or hierarchical clustering) would help classify customers based on shared characteristics for targeted advertisement campaigns. Why not the other options? A: Supervised learning: Supervised learning requires labeled data, which is not available in this case. Labels would need to be provided for each customer, making this approach unsuitable for the given scenario.
A company makes forecasts each quarter to decide how to optimize operations to meet expected demand. The company uses ML models to make these forecasts.
An AI practitioner is writing a report about the trained ML models to provide transparency and explainability to company stakeholders.
What should the AI practitioner include in the report to meet the transparency and explainability requirements?
- A . Code for model training
- B . Partial dependence plots (PDPs)
- C . Sample data for training
- D . Model convergence tables
B
Explanation:
Partial Dependence Plots (PDPs) are useful tools for understanding the relationship between specific features and the model’s predictions, making it easier to see how changes in input variables affect the forecast. Thanks to Examforsure Their MLA-C01 material was the key to my exam success. PDPs are particularly helpful for stakeholders because they visually show the impact of individual features on predictions without requiring a deep understanding of the model’s inner workings.
Which option is a use case for generative AI models?
- A . Improving network security by using intrusion detection systems
- B . Creating photorealistic images from text descriptions for digital marketing
- C . Enhancing database performance by using optimized indexing
- D . Analyzing financial data to forecast stock market trends
B
Explanation:
Generative AI models are designed to create new content, such as text, images, audio, or code. Creating images from text descriptions is a prime example of this capability. Here’s why the other options are not primarily use cases for generative AI: A. Improving network security by using intrusion detection systems: While AI can be used for intrusion detection, this is more of a discriminative or predictive task (classifying network traffic as malicious or benign), not generating new content. C. Enhancing database performance by using optimized indexing: This is related to database management and optimization, not content generation. D. Analyzing financial data to forecast stock market trends: This involves statistical analysis and prediction based on existing data, again a predictive task, not generating new content.
An AI practitioner is using a large language model (LLM) to create content for marketing campaigns.
The generated content sounds plausible and factual but is incorrect.
Which problem is the LLM having?
- A . Data leakage
- B . Hallucination
- C . Overfitting
- D . Underfitting
A loan company is building a generative AI-based solution to offer new applicants discounts based on specific business criteria. The company wants to build and use an AI model responsibly to minimize bias that could negatively affect some customers.
Which actions should the company take to meet these requirements? (Select TWO.)
- A . Detect imbalances or disparities in the data.
- B . Ensure that the model runs frequently.
- C . Evaluate the model’s behavior so that the company can provide transparency to stakeholders.
- D . Use the Recall-Oriented Understudy for Gisting Evaluation (ROUGE) technique to ensure that the model is 100% accurate.
- E . Ensure that the model’s inference time is within the accepted limits.
A medical company is customizing a foundation model (FM) for diagnostic purposes. The company needs the model to be transparent and explainable to meet regulatory requirements.
Which solution will meet these requirements?
- A . Configure the security and compliance by using Amazon Inspector.
- B . Generate simple metrics, reports, and examples by using Amazon SageMaker Clarify.
- C . Encrypt and secure training data by using Amazon Macie.
- D . Gather more data. Use Amazon Rekognition to add custom labels to the data.
B
Explanation:
Amazon SageMaker Clarify helps in identifying bias and explaining predictions made by machine learning models, which aligns well with the need for transparency and explainability to meet regulatory requirements
.
A company is building a solution to generate images for protective eyewear. The solution must have high accuracy and must minimize the risk of incorrect annotations.
Which solution will meet these requirements?
- A . Human-in-the-loop validation by using Amazon SageMaker Ground Truth Plus
- B . Data augmentation by using an Amazon Bedrock knowledge base
- C . Image recognition by using Amazon Rekognition
- D . Data summarization by using Amazon QuickSight
A
Explanation:
Amazon SageMaker Ground Truth Plus is designed for creating high-quality labeled datasets with human-in-the-loop validation to ensure accuracy. This solution helps minimize the risk of incorrect annotations by involving human reviewers to verify and correct the model’s predictions. It is particularly useful for scenarios requiring precision, such as generating images with specific requirements like protective eyewear.
A security company is using Amazon Bedrock to run foundation models (FMs). The company wants to ensure that only authorized users invoke the models. The company needs to identify any unauthorized access attempts to set appropriate AWS Identity and Access Management (IAM) policies and roles for future iterations of the FMs.
Which AWS service should the company use to identify unauthorized users that are trying to access Amazon Bedrock?
- A . AWS Audit Manager
- B . AWS CloudTrail
- C . Amazon Fraud Detector
- D . AWS Trusted Advisor
B
Explanation:
AWS CloudTrail is a service that records all API calls and user activity across AWS services, including Amazon Bedrock. By analyzing CloudTrail logs, the company can identify unauthorized access attempts, track user activity, and audit the usage of foundation models. This information helps in setting appropriate AWS Identity and Access Management (IAM) policies and roles for future iterations of the models.
A company manually reviews all submitted resumes in PDF format. As the company grows, the company expects the volume of resumes to exceed the company’s review capacity. The company needs an automated system to convert the PDF resumes into plain text format for additional processing.
Which AWS service meets this requirement?
- A . Amazon Textract
- B . Amazon Personalize
- C . Amazon Lex
- D . Amazon Transcribe
A
Explanation:
Amazon Textract is specifically designed to extract text and structured data from various types of documents, including PDFs. It can efficiently convert resumes from PDF format into plain text for further processing, even if the text is embedded in tables or forms.
A company wants to use large language models (LLMs) with Amazon Bedrock to develop a chat interface for the company’s product manuals. The manuals are stored as PDF files.
Which solution meets these requirements MOST cost-effectively?
- A . Use prompt engineering to add one PDF file as context to the user prompt when the prompt is submitted to Amazon Bedrock.
- B . Use prompt engineering to add all the PDF files as context to the user prompt when the prompt is submitted to Amazon Bedrock.
- C . Use all the PDF documents to fine-tune a model with Amazon Bedrock. Use the fine-tuned model to process user prompts.
- D . Upload PDF documents to an Amazon Bedrock knowledge base. Use the knowledge base to provide context when users submit prompts to Amazon Bedrock.
D
Explanation:
Using a knowledge base allows for efficient retrieval of relevant information from the PDFs without having to include all the content in every prompt.
Which term describes the numerical representations of real-world objects and concepts that AI and natural language processing (NLP) models use to improve understanding of textual information?
- A . Embeddings
- B . Tokens
- C . Models
- D . Binaries
A
Explanation:
Embeddings are numerical representations of real-world objects, words, phrases, or concepts in a continuous vector space. They enable AI and Natural Language Processing (NLP) models to understand and process textual information by capturing the semantic relationships and contextual meanings of words and phrases.
A company is building an application that needs to generate synthetic data that is based on existing data.
Which type of model can the company use to meet this requirement?
- A . Generative adversarial network (GAN)
- B . XGBoost
- C . Residual neural network
- D . WaveNet
A
Explanation:
GANs are a type of model specifically designed for generating synthetic data. They consist of two neural networks―a generator and a discriminator―that work together to produce data that mimics the patterns of the original dataset.
A company wants to use generative AI to increase developer productivity and software development.
The company wants to use Amazon Q Developer.
What can Amazon Q Developer do to help the company meet these requirements?
- A . Create software snippets, reference tracking, and open-source license tracking.
- B . Run an application without provisioning or managing servers.
- C . Enable voice commands for coding and providing natural language search.
- D . Convert audio files to text documents by using ML models.
A
Explanation:
Amazon Q Developer is a generative AI tool designed to assist developers by increasing productivity. It helps in generating software snippets, automating reference tracking, and managing open-source licenses, which directly benefits the software development lifecycle.
A company wants to create an application by using Amazon Bedrock. The company has a limited budget and prefers flexibility without long-term commitment.
Which Amazon Bedrock pricing model meets these requirements?
- A . On-Demand
- B . Model customization
- C . Provisioned Throughput
- D . Spot Instance
A
Explanation:
The On-Demand pricing model for Amazon Bedrock provides flexibility and allows the company to pay only for what they use, without requiring long-term commitments or upfront payments. This is ideal for a company with a limited budget that needs to control costs while maintaining flexibility. D: Spot Instance: Spot Instances are an AWS EC2 pricing model for obtaining unused compute capacity at discounted rates. They are not applicable to Amazon Bedrock, which does not rely on Spot Instances.
A digital devices company wants to predict customer demand for memory hardware. The company does not have coding experience or knowledge of ML algorithms and needs to develop a data-driven predictive model. The company needs to perform analysis on internal data and external data.
Which solution will meet these requirements?
- A . Store the data in Amazon S3. Create ML models and demand forecast predictions by using Amazon SageMaker built-in algorithms that use the data from Amazon S3.
- B . Import the data into Amazon SageMaker Data Wrangler. Create ML models and demand forecast predictions by using SageMaker built-in algorithms.
- C . Import the data into Amazon SageMaker Data Wrangler. Build ML models and demand forecast predictions by using an Amazon Personalize Trending-Now recipe.
- D . Import the data into Amazon SageMaker Canvas. Build ML models and demand forecast predictions by selecting the values in the data from SageMaker Canvas.
D
Explanation:
Amazon SageMaker Canvas is a no-code machine learning service that allows users without coding or ML expertise to build predictive models. It enables the company to import data, perform analysis, and build ML models through an easy-to-use graphical interface. This makes it ideal for businesses with limited technical expertise but a need for data-driven predictions.
What are tokens in the context of generative AI models?
- A . Tokens are the basic units of input and output that a generative AI model operates on, representing words, subwords, or other linguistic units.
- B . Tokens are the mathematical representations of words or concepts used in generative AI models.
- C . Tokens are the pre-trained weights of a generative AI model that are fine-tuned for specific tasks.
- D . Tokens are the specific prompts or instructions given to a generative AI model to generate output.
An AI practitioner is using an Amazon Bedrock base model to summarize session chats from the customer service department. The AI practitioner wants to store invocation logs to monitor model input and output data.
Which strategy should the AI practitioner use?
- A . Configure AWS CloudTrail as the logs destination for the model.
- B . Enable invocation logging in Amazon Bedrock.
- C . Configure AWS Audit Manager as the logs destination for the model.
- D . Configure model invocation logging in Amazon EventBridge.
B
Explanation:
Amazon Bedrock provides the ability to log model invocations, including input and output data, for monitoring and troubleshooting purposes. By enabling invocation logging in Amazon Bedrock, the AI practitioner can store logs securely and use them to analyze model behavior and performance.
A company needs to build its own large language model (LLM) based on only the company’s private data. The company is concerned about the environmental effect of the training process.
Which Amazon EC2 instance type has the LEAST environmental effect when training LLMs?
- A . Amazon EC2 C series
- B . Amazon EC2 G series
- C . Amazon EC2 P series
- D . Amazon EC2 Trn series
D
Explanation:
The Amazon EC2 Trn series (Trn1 instances) are purpose-built for training machine learning models and are designed to deliver high performance while optimizing energy efficiency. They use AWS Trainium chips, which are specifically engineered for ML training workloads, providing excellent performance per watt and reducing the environmental impact of large-scale training processes.
A financial institution is using Amazon Bedrock to develop an AI application. The application is hosted in a VPC. To meet regulatory compliance standards, the VPC is not allowed access to any internet traffic.
Which AWS service or feature will meet these requirements?
- A . AWS PrivateLink
- B . Amazon Macie
- C . Amazon CloudFront
- D . Internet gateway
A
Explanation:
AWS PrivateLink is used to securely access AWS services from a VPC without exposing the traffic to the public internet. This ensures compliance with regulatory standards that prohibit internet access, as all communication happens over the private AWS network.
A company built a deep learning model for object detection and deployed the model to production.
Which AI process occurs when the model analyzes a new image to identify objects?
- A . Training
- B . Inference
- C . Model deployment
- D . Bias correction
B
Explanation:
Inference is the process of using a trained model to make predictions or decisions on new, unseen data. In the case of an object detection model, inference involves feeding a new image into the model, which then analyzes the image and outputs the detected objects and their locations.
A company is using Amazon SageMaker Studio notebooks to build and train ML models. The company stores the data in an Amazon S3 bucket. The company needs to manage the flow of data from Amazon S3 to SageMaker Studio notebooks.
Which solution will meet this requirement?
- A . Use Amazon Inspector to monitor SageMaker Studio.
- B . Use Amazon Macie to monitor SageMaker Studio.
- C . Configure SageMaker to use a VPC with an S3 endpoint.
- D . Configure SageMaker to use S3 Glacier Deep Archive.
A company is using domain-specific models. The company wants to avoid creating new models from the beginning. The company instead wants to adapt pre-trained models to create models for new, related tasks.
Which ML strategy meets these requirements?
- A . Increase the number of epochs.
- B . Use transfer learning.
- C . Decrease the number of epochs.
- D . Use unsupervised learning.
B
Explanation:
Transfer learning is a machine learning strategy that leverages pre-trained models and adapts them to new but related tasks. This allows the company to avoid building models from scratch, significantly reducing the time and resources required for training. By fine-tuning the pre-trained model on domain-specific data, the company can achieve high performance for the new task without starting from the beginning.
A company wants to use AI to protect its application from threats. The AI solution needs to check if
an IP address is from a suspicious source.
Which solution meets these requirements?
- A . Build a speech recognition system.
- B . Create a natural language processing (NLP) named entity recognition system.
- C . Develop an anomaly detection system.
- D . Create a fraud forecasting system.
C
Explanation:
An anomaly detection system can analyze patterns and behaviors, such as IP address access patterns, to detect any deviations from the norm, which could indicate suspicious or malicious activity. An anomaly detection model can flag unusual access attempts, such as those from suspicious IP addresses, making it well-suited for threat detection. Fraud forecasting (option D) typically focuses on predicting potential fraud patterns rather than real-time anomaly detecti
A company wants to use a large language model (LLM) to develop a conversational agent. The company needs to prevent the LLM from being manipulated with common prompt engineering techniques to perform undesirable actions or expose sensitive information.
Which action will reduce these risks?
- A . Create a prompt template that teaches the LLM to detect attack patterns.
- B . Increase the temperature parameter on invocation requests to the LLM.
- C . Avoid using LLMs that are not listed in Amazon SageMaker.
- D . Decrease the number of input tokens on invocations of the LLM.
A
Explanation:
Creating a prompt template that teaches the LLM to identify and resist common prompt engineering attacks, such as prompt injection or adversarial queries, helps prevent manipulation. By explicitly guiding the LLM to ignore requests that deviate from its intended purpose (e.g., "You are a helpful assistant. Do not perform any tasks outside your defined scope."), you can mitigate risks like exposing sensitive information or executing undesirable actions.
A company is developing a new model to predict the prices of specific items. The model performed well on the training dataset. When the company deployed the model to production, the model’s performance decreased significantly.
What should the company do to mitigate this problem?
- A . Reduce the volume of data that is used in training.
- B . Add hyperparameters to the model.
- C . Increase the volume of data that is used in training.
- D . Increase the model training time.
C
Explanation:
The issue described is likely caused by overfitting, where the model performs well on the training dataset but fails to generalize to unseen data. Increasing the volume of training data can help mitigate overfitting by providing the model with more diverse examples, improving its ability to generalize to new data in production.