AWS Certified AI Practitioner (AIF-C01): Complete Study Guide

The AWS Certified AI Practitioner (AIF-C01) validates foundational knowledge of AI, ML, and generative AI on AWS. This guide covers all five domains — strategies, service comparisons, traps, and practice quizzes.

Exam Overview#

---

Domain Breakdown#

---

Domain 1: Fundamentals of AI and ML (20%)#

This domain tests your understanding of core machine learning concepts and the AWS AI/ML service stack.

Types of Machine Learning#

Classification vs Regression#

AWS AI/ML Service Stack#

AWS offers AI/ML services at three levels:

┌─────────────────────────────────────────────────────────┐
│  AI SERVICES (Pre-built, No ML expertise needed)        │
│  Rekognition, Textract, Comprehend, Transcribe, Polly   │
├─────────────────────────────────────────────────────────┤
│  ML SERVICES (Managed infrastructure)                    │
│  Amazon SageMaker, Bedrock, Amazon Q                     │
├─────────────────────────────────────────────────────────┤
│  ML FRAMEWORKS (Full control)                            │
│  TensorFlow, PyTorch on EC2/ECS                          │
└─────────────────────────────────────────────────────────┘

Key AWS AI Services#

---

Domain 2: Fundamentals of Generative AI (24%)#

This domain covers foundation models, LLMs, prompt engineering, and techniques like RAG and fine-tuning.

What Are Foundation Models?#

Foundation models are large AI models trained on vast datasets that can be adapted for many downstream tasks. Key characteristics:

• Pre-trained on massive datasets (internet text, images, code)
• General-purpose — can be adapted for specific tasks
• Transfer learning — knowledge transfers to new domains

Large Language Models (LLMs)#

LLMs are foundation models specifically trained on text data. They can:

• Generate human-like text
• Answer questions
• Summarize documents
• Translate languages
• Write and explain code

Key Inference Parameters#

Prompt Engineering Best Practices#

RAG vs Fine-Tuning#

---

Domain 3: Applications of Foundation Models (28%)#

This is the highest-weighted domain. It covers Amazon Bedrock, Amazon Q, and building AI applications.

Amazon Bedrock Overview#

Amazon Bedrock is a fully managed service for accessing foundation models via API.

Foundation Models in Bedrock#

Amazon Titan Models#

Amazon Bedrock Knowledge Bases#

Knowledge Bases enable RAG by:

1. Ingesting your documents (S3, web crawl, etc.)
2. Chunking and embedding documents
3. Storing vectors in a vector database
4. Retrieving relevant chunks at query time

Amazon Bedrock Agents#

Agents can:

• Execute multi-step tasks
• Call external APIs (action groups)
• Access knowledge bases
• Maintain conversation context

Amazon Q#

---

Domain 4: Guidelines for Responsible AI (14%)#

This domain covers bias, fairness, explainability, and content moderation.

Types of Bias in ML#

AWS Tools for Responsible AI#

Model Explainability#

Why it matters:

• Regulatory compliance (especially financial services, healthcare)
• Building user trust
• Debugging model behavior
• Identifying bias

SageMaker Clarify provides:

• Feature importance rankings
• SHAP (SHapley Additive exPlanations) values
• Partial dependence plots
• Bias metrics across groups

Bedrock Guardrails#

Guardrails help you implement responsible AI policies:

---

Domain 5: Security, Compliance, and Governance (14%)#

This domain covers data privacy, VPC configurations, model governance, and compliance.

Data Privacy in AI Workloads#

VPC Endpoints for AI Services#

AWS PrivateLink enables private connectivity:

• Traffic stays within AWS network
• No internet gateway required
• Enhanced security posture

Model Governance with SageMaker#

Compliance Considerations#

---

AWS AI Services Quick Reference#

---

Common Exam Traps to Avoid#

1. RAG vs Fine-Tuning#

• RAG: Use when you need citations, current information, or specific knowledge
• Fine-tuning: Use when you need to change model behavior or style

2. Temperature vs Max Tokens#

• Temperature: Controls creativity/randomness
• Max tokens: Controls response length

3. Embeddings vs Text Generation#

• Embeddings (Titan Embeddings): Convert text to vectors for search
• Text generation (Titan Text): Generate new text

4. VPC Endpoints vs Encryption#

• VPC endpoints: Control WHERE traffic goes (network path)
• Encryption: Protect data content (at rest or in transit)

5. Clarify vs Guardrails#

• Clarify: Bias detection and explainability for YOUR models
• Guardrails: Content filtering for Bedrock inputs/outputs

---

4-Week Study Plan#

Week 1: Foundations#

• Complete Domain 1 (AI/ML Fundamentals)
• Complete Domain 2 (Generative AI)
• Hands-on: Try Amazon Bedrock playground

Week 2: Applications#

• Complete Domain 3 (Foundation Models)
• Hands-on: Build a simple RAG application
• Hands-on: Test different models in Bedrock

Week 3: Governance#

• Complete Domain 4 (Responsible AI)
• Complete Domain 5 (Security)
• Hands-on: Configure Guardrails

Week 4: Review and Practice#

• Take practice exams
• Review weak areas
• Focus on service comparisons

---

Know your range before test day#

You've worked the domain quizzes. Sit a short AIF-C01 diagnostic next — then decide whether free practice or a timed sitting is the right move.

• Timed diagnostic with domain-level breakdown
• Clear signal on what to drill next
• Practice free; timed sittings and readiness range on Pro

Take the free AI Practitioner diagnostic →

---

Next step: Start practicing free on the AWS AI Practitioner (AIF-C01) hub, or sit a short diagnostic first.