Fundamentals of ML and AIFull notesSummaryRingkasanStoriesPracticeAWS Infrastructure and Technologies18 exam-style questions on this lesson.Question 1 of 18A data science team wants a fully managed service to build, train, and deploy its own custom ML models. Which service fits?AAmazon SageMaker AIBPartyRockCAmazon ComprehendDAmazon PollyCheck answerQuestion 2 of 18A company wants to analyze thousands of customer emails to find the overall sentiment and the key phrases in each one. It has no ML expertise. Which service fits?AAmazon RekognitionBAmazon ComprehendCAmazon TranslateDAmazon TextractCheck answerQuestion 3 of 18An insurance company receives scanned claim forms. It must pull out the form fields and tables, keeping each value linked to its field. Which service fits?AAmazon ComprehendBAmazon KendraCAmazon TextractDAmazon TranscribeCheck answerQuestion 4 of 18A company wants a voice bot on its phone line that understands what callers say and routes them. Which service provides speech recognition and natural language understanding for chatbots?AAmazon PollyBAmazon TranslateCAmazon PersonalizeDAmazon LexCheck answerQuestion 5 of 18A news website wants to offer audio versions of its articles read in a natural-sounding voice. Which service fits?AAmazon PollyBAmazon ComprehendCAmazon LexDAmazon TranscribeCheck answerQuestion 6 of 18A company wants to create searchable text records of recorded meetings, with a timestamp for every word. Which service fits?AAmazon KendraBAmazon TranscribeCAmazon TextractDAmazon PollyCheck answerQuestion 7 of 18A social media platform must detect inappropriate content in user-uploaded images and videos. Which service fits?AAmazon PersonalizeBAmazon TextractCAmazon RekognitionDAmazon ComprehendCheck answerQuestion 8 of 18Employees need to search across the company's wikis, file shares, and document repositories using natural-language questions. Which service fits?AAmazon LexBAmazon TranslateCAmazon PersonalizeDAmazon KendraCheck answerQuestion 9 of 18An online store wants to show each shopper real-time product recommendations based on their browsing and purchase history. Which service fits with the least ML expertise?AAmazon PersonalizeBAmazon ComprehendCAWS DeepRacerDAmazon KendraCheck answerQuestion 10 of 18A team wants a hands-on way to learn reinforcement learning. Which AWS offering uses a 1/18th scale autonomous race car?AAmazon Q DeveloperBAWS DeepRacerCPartyRockDAmazon SageMaker JumpStartCheck answerQuestion 11 of 18A developer wants one-click deployment and fine-tuning of popular open-source models inside SageMaker. Which feature fits?AAmazon KendraBPartyRockCAmazon SageMaker JumpStartDAmazon BedrockCheck answerQuestion 12 of 18A company wants to build a generative AI app using foundation models from several providers through a single API, without managing any servers, and customize the models privately with its own data. Which service fits?AAWS DeepRacerBAmazon EC2 with GPUsCAmazon SageMaker AI training jobsDAmazon BedrockCheck answerQuestion 13 of 18A marketing employee with no coding experience wants to experiment with building simple generative AI apps. Which offering is an Amazon Bedrock playground for this?APartyRockBAmazon SageMaker StudioCAmazon Q DeveloperDAWS DeepRacerCheck answerQuestion 14 of 18Developers want generative AI code suggestions directly inside their IDE. Which service fits?AAmazon ComprehendBAmazon Q DeveloperCAmazon LexDAmazon Q BusinessCheck answerQuestion 15 of 18. Select two.Which TWO are benefits of building AI solutions on AWS? (Select TWO.)AManaged services that handle infrastructure, training, and deploymentBPay-as-you-go pricing with global infrastructure that scalesCEvery model is free to useDNo need for any dataEAWS takes full responsibility for securing customer dataCheck answer0 of 2 selectedQuestion 16 of 18A company uses Amazon Bedrock and wants its bill to track exactly how much text its application sends and receives. How does Bedrock on-demand pricing work?ACharges per user accountBA fixed monthly fee regardless of useCCharges per token processed or generatedDCharges per GPU purchasedCheck answerQuestion 17 of 18A company's generative AI app has steady, heavy traffic and needs predictable performance. It is willing to pay up front for reserved capacity. Which option fits?AA single Availability Zone deploymentBOn-demand token pricingCCPU-only instancesDProvisioned throughputCheck answerQuestion 18 of 18Which change will usually increase the cost of an AI deployment?ADeploying across multiple Availability Zones and Regions for redundancyBShortening the prompts the application sendsCServing all users from fewer AWS RegionsDMoving a light workload from GPUs to CPU instancesCheck answerGenerative AI Fundamentals16 exam-style questions on this lesson.Artificial Intelligence4 exam-style questions on this lesson.