Skip to content

06-aws-services-primer.md

Latest commit

 

History

History
80 lines (60 loc) · 5.22 KB

06-aws-services-primer.md

File metadata and controls

80 lines (60 loc) · 5.22 KB
Raw

AWS services primer (exam-oriented)

This primer ties in-scope services from the official in-scope list to exam-style mental models. Use AWS documentation for details; this page is a shortcut, not a spec.

Machine Learning (core names)

Service One-line role
Amazon Bedrock Managed foundation models, knowledge bases, agents, guardrails.
Amazon SageMaker AI Build, train, deploy, monitor ML; JumpStart for pre-trained models.
Amazon Q Developer AI assistant for builders (IDE/in-console help).
Amazon Q Business Enterprise Q&A across connected company data (governed access).
Amazon Nova AWS foundation model family (multimodal capabilities vary by model).
Amazon Kendra Intelligent enterprise search (often paired with retrieval patterns).
Amazon Lex Build conversational bots.
Amazon Polly Text-to-speech.
Amazon Transcribe Speech-to-text.
Amazon Translate Machine translation.
Amazon Comprehend NLP insights (entities, sentiment, classification).
Amazon Rekognition Image and video analysis.
Amazon Textract Extract text and structured data from documents.
Amazon Personalize Real-time personalization and recommendations.
Amazon Fraud Detector Fraud scoring with ML templates.
Amazon Augmented AI (A2I) Human review workflows.

Analytics and data (common AI companions)

Service Role in AI stories
Amazon S3 Lake/landing zone for documents, images, audio; Bedrock ingestion source.
AWS Glue / DataBrew ETL and data prep for features and curated datasets.
Amazon Redshift Warehouse analytics; often a reporting layer.
Amazon OpenSearch Service Search + vector patterns for retrieval.
AWS Lake Formation Governed data lake permissions.

The official in-scope page is authoritative; AWS updates exam scope from time to time.

Security, identity, compliance

Service Typical exam angle
IAM Least privilege for humans and workloads invoking Bedrock/SageMaker.
KMS Encryption keys for data and model artifacts.
Secrets Manager Secrets rotation and access control.
Macie Sensitive data discovery.
CloudTrail Audit API usage.
Config Track configuration drift.
Audit Manager Compliance evidence.
Artifact Download attestations and agreements.

Operations and cost

Service Role
CloudWatch Metrics, logs, alarms for endpoints and applications.
Trusted Advisor Guidance on cost, security, fault tolerance.
Cost Explorer / Budgets Track model spend and set alerts.
Well-Architected Tool Review architectures against best practices (including sustainability).

How to study services for AIF-C01

  1. Match service → problem (for example Textract → forms; Transcribe → audio notes).
  2. Know where Bedrock fits vs SageMaker AI vs application services (Lex, Comprehend).
  3. Understand security building blocks (IAM, KMS, logging) as cross-cutting requirements.

Official link