Amazon SageMaker
Unified data, analytics, and AI platform
The next generation of Amazon SageMaker (since late 2024) brings data, analytics, and AI together in one place: Unified Studio, lakehouse, and data governance; the model build/train/deploy part is now called SageMaker AI.
- Cost$$ Medium
- Difficulty●●●●●
- Costs money when idle?A little
- Managed by AWS
- In examsSAA
Think of it asLike a shared office building: the data warehouse, analysis rooms, and AI workshop share one entrance.
When to use it
✓ Use it when
Organizations that want a shared environment for data engineers, analysts, and data scientists.
✗ Avoid it when
Only need to train/deploy ML models → SageMaker AI; only using ready-made foundation models → Bedrock.
Common uses
- Collaborate on data with Unified Studio
- Build a lakehouse on S3 and Redshift
- Govern and share data with controls
What does it cost?
- Pricing model: Charged by the component services used (compute, queries, storage, governance).
