AWS Cloud Practitioner Study Session Eight

January 02, 2026

I am taking the AWS Cloud Practitioner Exam in approximately two days and want to ensure I am prepared. This series will serve as non-exhaustive note taking for the information that I am internalizing as I go.

Study materials:

Notes:

AWS AI/ML stack is composed of 3 tiers.

(1) AI Services - Pre-built models that are already trained to perform specific functions
(2) ML Services - Customized approach with Amazon SageMaker AI where you build, train, deploy own ML models with fully managed infra
(3) ML Frameworks / Infra - Custom approach to building models using purpose-built chips that integrate with popular ML frameworks

What is an ML Framework? A software library with pre-built, optimized components.

Tier 1: Pre-built AWS AI Services

Language Services
Computer Vision and Search Services
Conversational AI and Personalization Services

Language Services - When you need to interpret text / speech and turn it into something meaningful. (TTS, STT)

Amazon Comprehend - Uses NLP to extract key insights from docs. Develops insights by recognizing key phrases, language, sentiment.
Use Cases: Content Classification, Customer Sentiment Analyisis, Compliance Monitoring

Amazon Polly - Converts text in to lifelike speech. Supports multiple languages, different genders, accents.

Use Cases: Virtual assistants, e-learning apps, accessibility enhancements for visually impaired users

Amazon Transcribe - Converts speech into text. Supports multiple languages. Features: Speaker identification, custom vocabulary, real-time transcription.

Use Cases: Customer call transcription, automated subtitling, metadata generation for media content

Amazon Translate - Text translation service that supports real-time and batch text translation across multiple languages

Use Cases: Document translation and multi-language application integrations

Amazon Kendra - UsesNLP to search for answers within enterprise content.

Use Cases: Intelligent search, chatbots, application search integration

Amazon Rekognition - Video analysis service. Can identify objects, people, text, scenes, activities within images and videos stored in Amazon S3.

Use Cases: Content moderation, identity verification, media analysis, home automation services

Amazon Textract - Detects and extracts typed and handwritten text found in documents, forms, tables within documents

Use Cases: Financial, healthcare, government form text extraction for quick processing

Amazon Lex - NLU and ASR to create lifelike conversations.

Use Cases: Virtual assistants, natural language search for FAQs, automated application bots

Amazon Personalize - Can use historical data to build intelligent applications with personalized customer recommendations

Use Cases: Personalized streaming, product, trending recommendations

Tier 2: ML Services

Provides a more customized approach for more control without having to manage infrastructure
SageMaker AI is a key offering in this tier

SageMaker AI: Fully managed service, can build, train, deploy ML models without worrying about infrastructure. IDE. Can track training, visualize data, debug workflows. Access to pre-trained models to deploy. Benefits - Choice of ML Tools: Increase innovation with different tools (IDE, no-code interface) Fully managed Infra: Focus on ML model development whle SageMaker AI provides with high-performance cost-effective infrastructure Repeatable ML Workflows: Automate / standardize MLOps practices and governance across your enterprise to support transparency and auditability

Introduction to Gen AI on AWS

Amazon SageMaker JumpStart - An ML hub with FMs and pre-built ML solutions deployable with a few clicks
Use Cases: Quickly deploy pre-trained models, fine-tune with domain-specific data, compare performance for different models
Amazon Bedrock - Fully managed service for adapting FMs from Amazon and elsewhere. Provides access to FMs through a single unified API.
Use Cases - Build gen AI apps, create apps that can generate multiple content types (multimodel), conversational agents
Amazon Q - Interactive AI assistant that can be integrated with company’s info repositories.
Amazon Q Business - Can answer pressing questions, help solve problems, take actions using data and expertise found in company’s info repositories Use Cases: Information requests, automated workflows, insight extract
Amazon Q Developer - Provides code recommendations to accelerate development for coding langauges. Use Cases: Faster code generation, improved reliability / security, automated code reviews
Data Pipelines for ETL Processes
(1) Extract data from arious sources and store it.
(2) Transform data into consistent, usable format for downstream tools to consume.
(3) Load it into destination system like data warehouse or analytics platform.
Data Analytics
Use Cases - Loan companies explaining lending decisions to customers, medical researchers analyzing clinical trial data through hypothesis testing, insurance companies making risk assessment models transparent to regulators.

AWS Data Pipeline Services

Amazon Kinesis Data Streams
Real-time ingestion of terabytes of data from applications, streams, sensors.
Automatic provisioning and scaling in on-demand mode.
Amazon Data Firehose
Data ingestion in near real-time. Provides automatic provisioning and scaling. Delivers data within seconds to data lakes, warehouses, analytics services.
Amazon S3 - Fully elastic, automatically scaling as you add / remove data.
Amazon Redshift - Fully manageed data warehouse service that can store petabytes of structured / semistructured data.

Data Cataloging Services
AWS Glue Data Catalog - Centralized, scalable, managed metadata repository that enhances data discovery.

Data Processing Services
AWS Glue - Fully managed ETL, makes data prep simpler, faster, cost effective. Best suited for data processing in data pipeline.
Amazon EMR - Automatically handles infra provisioning, cluster management, scaling. Supports Apache Spark, Apache Hadoop, Apache Hive.

Data Analysis and Visualization Services
Amazon Athena - Serverless service that can access data hosted on Amazon S3, on-prem, or multicloud and runs SQL queries to analyze data in relational, non-relational, object, custom data sources.

Amazon Redshift - Fully managed data warehouse solution, columnar storage, massively parallel processesing architecture makes it ideal for analyzing large datasets.
Can use it to perform SQL queries on large datasets for frequent, high-performance analytics workloads

Amazon QuickSight - Technical and non-technical users can quickly create modern interactive dashboards / reports from data sources without managing infra.
Natural language queries
Amazon OpenSearch Service - Can search for relevant content through precise keyword matching or natural language queries.
Can use OpenSearch Service to visualize data in a data pipeline.
Amazon S3 can store virtually unlimited amounts of unstructured data. This makes it a popular data lake choice and best option for the team.