Basic Information
| Field | Details |
|---|---|
| English name | AWS AI/ML Stack |
| Full name | AWS AI Services / AWS ML Services / ML Frameworks and Infrastructure |
| Chinese description | AWS AI/ML 三层技术栈 |
| Japanese description | AWS AI/ML 3層スタック |
| Exam frequency | ⭐⭐⭐⭐⭐ |
| Often confused with | AI Services / SageMaker / EC2 + ML Frameworks |
In one sentence
Choose among pretrained AI services, ML platforms, and ML infrastructure based on control and operational burden.
Key points
- AI Services provide ready-made APIs such as Polly, Transcribe, Comprehend, Textract, and Rekognition.
- SageMaker provides the lifecycle for custom-model preparation, training, deployment, and monitoring.
- EC2, EKS, GPUs, and ML frameworks provide the most control and the most infrastructure responsibility.
- Bedrock serves generative-AI application development through managed foundation-model APIs.
Exam focus
- The more a question emphasizes ready capability, the higher the layer; custom training and control push lower.
Common pitfalls
- Do not choose by product name alone; confirm data type, latency, control, operations, and cost constraints.
Remember
AI APIs → SageMaker platform → self-managed ML infrastructure; control and operations both increase.