AWS Services Reference

AI & Machine Learning

Amazon Bedrock Managed foundation models (Claude, Llama, Mistral, Titan) with Knowledge Bases, Agents, Guardrails.
Amazon SageMaker End-to-end ML platform — train, tune, deploy, monitor custom models.

Amazon Comprehend NLP — entities, sentiment, PII redaction, medical concept extraction.
Amazon Textract Document AI — OCR, forms, tables, invoice/ID/lending extraction.

Amazon Rekognition Computer vision — labels, faces, moderation, custom image classifiers.
AI Services Overview How Bedrock, task-specific APIs, and SageMaker fit together.

Compute

Amazon EC2 Virtual machines — instance families, Spot, Savings Plans, Auto Scaling.
AWS Lambda Event-driven serverless functions with pay-per-invocation pricing.

ECS, EKS & Fargate Containers — proprietary orchestrator (ECS), Kubernetes (EKS), or serverless (Fargate).
AWS EMR Managed Hadoop, Spark, Hive, Presto, and Trino clusters.

Auto Scaling Scale EC2 and EMR processing resources up or down based on load.
AWS Step Functions State-machine orchestration for multi-step serverless workflows.

Database & Data Warehouse

Amazon RDS Managed relational DBs — MySQL, PostgreSQL, MariaDB, Oracle, SQL Server.
Amazon Aurora Cloud-native MySQL/Postgres engine with distributed storage & Global Database.

Amazon DynamoDB Serverless NoSQL key-value / document DB at any scale.
Amazon ElastiCache In-memory cache — Valkey, Redis OSS, Memcached.

Amazon Redshift Cloud data warehouse — MPP columnar SQL for analytics.
Data Mesh & Data Fabric Data architecture patterns for large organizations.

Networking & Content Delivery

Amazon VPC Isolated virtual networks — subnets, route tables, security groups, endpoints.

Amazon API Gateway Managed REST, HTTP, and WebSocket APIs.
Amazon CloudFront Global CDN with edge compute (CloudFront Functions, Lambda@Edge).

Amazon Route 53 Managed DNS, domain registration, health checks, traffic routing.

Storage

Amazon S3 Object storage — data lake foundation, storage classes, lifecycle rules.

S3 Transfer Acceleration Accelerated uploads over the CloudFront edge network.

Building an ETL Pipeline on AWS

Data Ingestion (Extract)

Amazon S3 Store raw data in S3 buckets — the foundation of AWS data lakes.

Kinesis Data Streams Durable, replayable real-time stream with multiple consumers.

Data Firehose Serverless streaming delivery to S3, Redshift, OpenSearch, Snowflake.

Data Transformation

AWS Glue (ETL) Managed serverless Spark for ETL — clean, format, enrich.

AWS Lambda Simple transformations in real-time or small batches.

AWS EMR Large-scale Spark/Hadoop for complex transformations.

Query & Load

Amazon Athena Serverless SQL directly over S3 — Parquet, Iceberg, Hudi.

Amazon Redshift Load transformed data for analytical queries and BI reporting.

Amazon QuickSight Native BI and dashboards with generative-AI Q.

Orchestration & Metadata

Glue Data Catalog Central metadata store for tables across Athena, Redshift Spectrum, EMR.
Lake Formation Fine-grained access control over S3-based data lakes built on Glue Catalog.
Lake vs Cloud Formation Side-by-side reference for two services with confusingly similar names.

AWS Glue Workflow Orchestrate Glue crawlers, jobs, and triggers as a DAG.

Step Functions General-purpose state machines across any AWS service.

Security, Monitoring & Governance

Security & Compliance

IAM Identities, roles, and policies controlling access to every AWS resource.
Secrets Manager Store, rotate, and retrieve credentials and API keys.

KMS Encryption keys and cryptographic operations across AWS.
AWS CloudTrail Audit log of every API call in your accounts.

Config & Inspector Resource configuration tracking and vulnerability scanning.

Monitoring

AWS CloudWatch Metrics, logs, alarms, dashboards, and anomaly detection.

CloudWatch Events Event-driven triggers (now EventBridge) for schedules and state changes.

CloudFormation Declarative infrastructure-as-code templates.