ExamPark
Course menuCLF-C02

Exams / CLF-C02 / Study guide / AI, machine learning and analytics

AI, machine learning and analytics: CLF-C02 study notes and practice questions

What to know about AI, machine learning and analytics for the CLF-C02 exam, from the Cloud Technology and Services domain (34% of the exam guide), then 3 free questions to test it. Part of the CLF-C02 study guide.

Last verified against current AWS documentation ·How we write our questions

What to know for the exam

This topic is pure matching: a business need on one side, an AWS service on the other.

AI and machine learning services

Need Service
Analyze images and video (faces, objects, unsafe content) Amazon Rekognition
Find sentiment, key phrases or entities in text Amazon Comprehend
Convert speech to text Amazon Transcribe
Convert text to lifelike speech Amazon Polly
Translate between languages Amazon Translate
Extract text and data from scanned documents and forms Amazon Textract
Build chatbots with voice and text Amazon Lex
Intelligent search across company documents Amazon Kendra
Product recommendations Amazon Personalize
Detect online fraud Amazon Fraud Detector
Build, train and deploy your own ML models Amazon SageMaker AI
Use foundation models to build generative AI applications Amazon Bedrock
A generative AI assistant for work and for developers Amazon Q

Analytics services

Need Service
SQL queries directly on data in S3, no servers Amazon Athena
Data warehouse for large-scale SQL analytics Amazon Redshift
Prepare and transform data (ETL), catalog it AWS Glue
Collect and process streaming data in real time Amazon Kinesis (Data Streams), Amazon Data Firehose to deliver it
Dashboards and business intelligence Amazon QuickSight
Big data frameworks such as Apache Spark Amazon EMR
Search and log analytics Amazon OpenSearch Service
Managed Apache Kafka Amazon MSK

Common traps

  • Transcribe vs Polly. Transcribe turns speech into text; Polly turns text into speech.
  • Comprehend vs Textract. Textract reads text out of documents; Comprehend understands text you already have.
  • SageMaker AI for a common task. If a ready-made AI service does the job, it’s the lower-effort answer.
  • Athena vs Redshift. Occasional queries on S3 point to Athena; heavy, frequent analytics point to Redshift.

Try the 3 practice questions

Practice AI, machine learning and analyticsTake the free practice exam

Question 1

Medium · choose 2 ·Link to this question (question 1)

A bank is building a voice assistant for its phone helpline.

It must turn callers' spoken words into text, and read replies back to callers in a natural-sounding voice, using managed AI services with no model training.

Which TWO services should it use? (Choose two.)

Choose two answers.

  1. A. Amazon Transcribe, which converts speech into text
  2. B. Amazon Comprehend, which finds sentiment and key phrases in text
  3. C. Amazon Translate, which converts text between languages
  4. D. Amazon Rekognition, which analyzes images and video
  5. E. Amazon Polly, which converts text into lifelike speech
Show the answer and explanation

Answers: A, E

  • A. Right: Amazon Transcribe is a fully managed speech-to-text service, so it turns callers' speech into text without the bank building a model. Learn more: Amazon Transcribe (opens in a new tab)
  • B. Wrong: Comprehend analyzes existing text for sentiment, entities and key phrases. It does not convert speech to text or create spoken replies.
  • C. Wrong: Translate converts text between languages. It is useful for multilingual callers, but it does not turn speech into text or produce audio.
  • D. Wrong: Rekognition analyzes images and video. It has no role in converting speech or producing spoken audio.
  • E. Right: Amazon Polly turns text into lifelike speech, so it can read the assistant's replies aloud to callers. Learn more: Amazon Polly (opens in a new tab)

Explanation

Transcribe converts speech to text and Polly converts text to lifelike speech, which covers both halves of the voice assistant. Comprehend, Translate and Rekognition each work on text in another way, or on images and video, and none of them handles the speech conversion.

Study

Question 2

Easier ·Link to this question (question 2)

A retailer receives thousands of written product reviews each day.

It wants software to label each review as positive, negative, neutral or mixed, without training its own machine-learning model.

Which service should it use?

Choose one answer.

  1. A. Amazon Polly, which converts text into lifelike speech
  2. B. Amazon Textract, which extracts text from scanned documents and forms
  3. C. Amazon SageMaker AI, which builds and trains custom machine-learning models
  4. D. Amazon Comprehend, which detects sentiment in text
Show the answer and explanation

Answer: D

  • A. Wrong: Polly creates spoken audio from text. It does not judge whether a written review is positive or negative.
  • B. Wrong: Textract reads text from scanned documents and forms. The reviews are already text, so Textract does not judge their sentiment.
  • C. Wrong: SageMaker AI is for building and training your own models. The retailer does not want to train a model, so this adds work it does not need.
  • D. Right: Comprehend is a managed natural language service that returns positive, negative, neutral or mixed sentiment for each text, with no model for the retailer to build. Learn more: Sentiment analysis in Amazon Comprehend (opens in a new tab)

Explanation

Amazon Comprehend returns sentiment as positive, negative, neutral or mixed, using a managed model. Polly and Textract do other jobs, and SageMaker AI would require the retailer to build and train its own model.

Study

Question 3

Medium ·Link to this question (question 3)

A retail chain has 40 TB of sales data from five systems.

Analysts need to run complex SQL queries across all of it and connect their existing business intelligence tools, with AWS managing the infrastructure.

Which service should it choose?

Choose one answer.

  1. A. Amazon DynamoDB, a key-value and document database for fast lookups by key
  2. B. Amazon Redshift, a managed data warehouse for SQL analytics
  3. C. Amazon ElastiCache, an in-memory cache for fast repeated reads
  4. D. Amazon Kinesis Data Streams, which carries streams of records in real time
Show the answer and explanation

Answer: B

  • A. Wrong: DynamoDB is built for fast lookups by key. It is not designed for complex SQL analysis across many terabytes of data from several systems.
  • B. Right: Redshift is a fully managed cloud data warehouse for SQL analytics on large volumes of data. It works with BI tools such as Amazon QuickSight and Tableau. Learn more: Amazon Redshift (opens in a new tab)
  • C. Wrong: ElastiCache speeds up reads of data stored elsewhere. It does not hold the full sales data set or run analysis queries on it.
  • D. Wrong: Kinesis Data Streams moves streams of records as they arrive. It does not store the data for analysts to query with SQL.

Explanation

Redshift is AWS's managed data warehouse for SQL analytics across large data sets, and it supports BI tools. DynamoDB and ElastiCache serve fast lookups or cached reads, and Kinesis Data Streams moves records in real time rather than storing them for analysis.

Study

Want more on AI, machine learning and analytics? The 5 full-length practice sets have 12 more AI, machine learning and analytics questions, 325 in all, none of them in the free exam. One payment of $14.99, no subscription. See the practice sets

Study AI, machine learning and analytics

More Cloud Technology and Services topics