Module 1: Domain 1: Fundamentals of AI and ML
1
1.1 AI, ML, and Deep Learning 3
1.3 Data Types and Modalities 4
1.4 Training and Inference 5
1.5 Practical AI Applications 7
1.7 AWS Managed AI Services 9
1.9 Models, Production, and Metrics Module 2: Domain 2: Generative AI
1
2.1 Tokens, Embeddings, and Vectors 2
2.2 Transformers and Foundation Models 4
2.4 Pre-training, Fine-tuning, and RAG 5
2.5 Multimodal and Diffusion Models 6
2.6 Capabilities and Limitations 7
2.7 Hallucinations and Nondeterminism 9
2.9 Amazon Q and PartyRock 10
2.10 SageMaker JumpStart and Generative Cost Tradeoffs Module 3: Domain 3: Foundation Model Applications
1
3.1 Foundation Model Applications 2
3.2 Embeddings and Vector Stores 3
3.3 Retrieval Augmented Generation Design 5
3.5 Bedrock Knowledge Bases 6
3.6 Prompt Chaining and Orchestration 7
3.7 Evaluate Foundation Model Outputs 8
3.8 Cost, Latency, and Tokens in Production 9
3.9 Human Oversight and Feedback 10
3.10 Choosing a Foundation Model Application Path Module 4: Domain 4: Responsible AI
1
4.1 Features of Responsible AI 2
4.2 Guardrails for Amazon Bedrock 3
4.3 Responsible Model Selection 4
4.4 Legal Risks of Generative AI 5
4.5 Dataset Characteristics 7
4.7 Detect and Monitor Bias 8
4.8 Transparent and Explainable Models 9
4.9 SageMaker Model Cards and Licensing 10
4.10 Safety, Transparency, and Human-Centered Design Module 5: Domain 5: Security Compliance Governance
1
5.1 Shared Responsibility for AI Systems 2
5.2 IAM Roles, Policies, and Permissions 3
5.3 Encryption at Rest and in Transit 4
5.4 Amazon Macie and AWS PrivateLink 5
5.5 Data Lineage, Cataloging, and Source Citation 6
5.6 Secure Data Engineering and Privacy Technologies 7
5.7 Prompt Injection, Leakage, and Guardrails 8
5.8 Hallucination Detection and Grounding 9
5.9 ISO, SOC, Artifact, and Audit Manager 10
5.10 Governance Tools, Lifecycle, and the Scoping Matrix