AI and machine learning
The exam asks which service solves each task: understanding text, speech, images and documents, or training your own models. No model programming required.
SageMaker AI
Amazon SageMaker AI
A platform to build, train and deploy your own machine learning models.
Key points
- Covers the whole cycle: prepare data, train, tune and deploy.
- Managed infrastructure for training and inference.
- For data scientists who need custom models.
On the exam
"Build, train and deploy your own ML models": SageMaker AI.
The AWS generative AI assistant for work and development.
Key points
- Amazon Q Developer: helps write, test, debug and operate code and AWS resources.
- Answers questions about AWS in the console and the IDE.
- Generative AI with enterprise security and privacy controls.
On the exam
"A generative AI assistant that helps developers on AWS": Amazon Q Developer.
Rekognition
Amazon Rekognition
Image and video analysis: objects, faces, text and inappropriate content.
Key points
- Face detection and comparison.
- Content moderation.
- No ML experience required.
On the exam
Images and videos (faces, objects, moderation): Rekognition.
Textract
Amazon Textract
Extracts text, handwriting, tables and forms from scanned documents.
Key points
- Goes beyond OCR: understands the structure of the document.
- Uses: invoices, forms, identity documents.
- Results in a structured format.
On the exam
"Extract data from scanned forms and documents": Textract.
Comprehend
Amazon Comprehend
Natural language processing (NLP) to find meaning in text.
Key points
- Sentiment analysis.
- Extraction of entities, key phrases and language.
- Detection of personal data (PII) in text.
On the exam
"Find the sentiment of customer reviews": Comprehend.
Lex
Amazon Lex
Builds conversational interfaces (chatbots) for voice and text.
Key points
- The same conversational technology as Alexa.
- Integrates with Amazon Connect and Lambda.
- Recognizes intents and collects information from the user.
On the exam
A voice or text chatbot: Lex.
Polly
Amazon Polly
Turns text into lifelike speech (text-to-speech).
Key points
- Dozens of voices and languages.
- Natural-sounding neural voices.
- Uses: accessibility, article readers, customer service.
On the exam
Text to speech: Polly. Speech to text: Transcribe.
Transcribe
Amazon Transcribe
Turns speech into text (speech-to-text).
Key points
- Batch and real-time transcription.
- Subtitles, call and meeting transcription.
- Identifies who is speaking.
On the exam
Speech to text (subtitles, call transcription): Transcribe.
Translate
Amazon Translate
Neural machine translation of text between languages.
Key points
- Real-time and batch translation.
- Custom terminology.
- Uses: localizing websites and user-generated content.
On the exam
"Translate content into other languages": Translate.