AWS AI Practitioner (AIF-C01)

Basic AI, ML, and deep learning concepts, the ML development lifecycle, and appropriate use cases.

• AI vs machine learning vs deep learning terminology
• Supervised, unsupervised, and reinforcement learning
• Types of inferencing (batch vs real-time)
• Structured, semi-structured, and unstructured data
• The ML development lifecycle (data collection, training, evaluation, deployment, monitoring)
• Features, labels, training and validation data, overfitting and underfitting
• When AI/ML is and isn't an appropriate solution

Free Fundamentals of AI and ML practice questions →Fundamentals of AI and ML study guide →

Core generative AI concepts, foundation models, tokens and embeddings, and AWS generative AI services.

• Foundation models, LLMs, tokens, embeddings, prompts, and completions
• How generative AI differs from traditional ML
• Capabilities and limitations of generative AI (including hallucination)
• Amazon Bedrock and its model providers
• Amazon Q (Business and Developer)
• Amazon SageMaker JumpStart for foundation models
• Use cases: summarization, chatbots, code generation, image generation

Free Fundamentals of Generative AI practice questions →Fundamentals of Generative AI study guide →

Designing applications with foundation models: prompt engineering, RAG, fine-tuning, evaluation, and selecting models.

• Prompt engineering techniques (zero-shot, few-shot, chain-of-thought)
• Inference parameters (temperature, top-p, max tokens)
• Retrieval-Augmented Generation (RAG) and Amazon Bedrock Knowledge Bases
• Fine-tuning and continued pre-training vs prompt-based approaches
• Vector databases and embeddings for retrieval
• Amazon Bedrock Agents for multi-step tasks
• Evaluating foundation model performance (human and automated, benchmarks)
• Selecting a model by cost, latency, modality, and capability

Free Applications of Foundation Models practice questions →Applications of Foundation Models study guide →

Responsible AI dimensions, bias and fairness, transparency, and tools such as Amazon Bedrock Guardrails.

• Responsible AI dimensions: fairness, bias, robustness, transparency, explainability, safety
• Detecting and mitigating bias in data and models
• Amazon Bedrock Guardrails for content filtering and safety
• Amazon SageMaker Clarify for bias detection and explainability
• Model transparency: model cards and documentation
• Environmental and governance considerations of AI

Free Guidelines for Responsible AI practice questions →Guidelines for Responsible AI study guide →

Securing AI systems, data governance, access control, and compliance for AI workloads on AWS.

• Securing AI systems: IAM, encryption, least privilege, data protection
• Data governance, lineage, and privacy for AI workloads
• AWS shared responsibility model applied to AI/ML services
• Compliance standards and AWS services (AWS Artifact, AWS Audit Manager)
• Monitoring and logging AI systems (Amazon CloudWatch, AWS CloudTrail)
• Secure model and data access with Amazon Bedrock and SageMaker

Free Security, Compliance, and Governance for AI Solutions practice questions →Security, Compliance, and Governance for AI Solutions study guide →