1. Fundamentals of AI and ML - 20%
Explain basic AI concepts and terminologies
Define basic AI terms (for example, AI, ML, deep learning, neural networks, computer vision, natural language processing [NLP], model, algorithm, training and inferencing, bias, fairness, fit, large language models (LLMs))
Describe the similarities and differences between AI, ML, GenAI, and deep learning
Describe various types of inferencing (for example, batch, real-time)
Describe the different types of data in AI models (for example, labeled and unlabeled, tabular, time-series, image, text, structured and unstructured)
Describe supervised learning, unsupervised learning, and reinforcement learning
Identify practical use cases for AI
Recognize applications where AI/ML can provide value (for example, assist human decision making, solution scalability, automation)
Determine when AI/ML solutions are not appropriate (for example, cost-benefit analyses, situations when a specific outcome is needed instead of a prediction)
Select the appropriate ML techniques for specific use cases (for example, regression, classification, clustering)
Identify examples of real-world AI applications (for example, computer vision, NLP, speech recognition, recommendation systems, fraud detection, forecasting)
Explain the capabilities of AWS managed AI/ML services (for example, Amazon SageMaker AI, Amazon Transcribe, Amazon Translate, Amazon Comprehend, Amazon Lex, Amazon Polly)
Describe the ML development lifecycle
Describe components of an ML pipeline (for example, data collection, exploratory data analysis [EDA], data pre-processing, feature engineering, model training, hyperparameter tuning, evaluation, deployment, monitoring)
Describe sources of ML models (for example, open source pre-trained models, training custom models)
Describe methods to use a model in production (for example, managed API service, self-hosted API)
Identify relevant AWS services and features for each stage of an ML pipeline (for example, SageMaker AI, SageMaker Data Wrangler, SageMaker Feature Store, SageMaker Model Monitor)
Describe fundamental concepts of ML operations (MLOps) (for example, experimentation, repeatable processes, scalable systems, managing technical debt, achieving production readiness, model monitoring, model re-training)
Describe model performance metrics (for example, accuracy, AUC, F1 score) and business metrics (for example, cost per user, development costs, customer feedback, ROI)
2. Fundamentals of GenAI - 24%
Explain the basic concepts of GenAI
Define foundational GenAI concepts (for example, tokens, chunking, embeddings, vectors, prompt engineering, transformer-based LLMs, foundation models, multimodal models, diffusion models)
Identify potential use cases for GenAI models (for example, content generation, summarization, assistants, translation, code generation, customer service, search, recommendations)
Describe the foundation model lifecycle (for example, data selection, model selection, pre-training, fine-tuning, evaluation, deployment, feedback)
Understand the capabilities and limitations of GenAI
Capabilities and limitations
Describe advantages (for example, adaptability, responsiveness, simplicity)
Identify disadvantages (for example, hallucinations, interpretability, inaccuracy, nondeterminism)
Identify model selection factors (for example, performance, constraints, compliance)
Determine business value and metrics (for example, efficiency, conversion rate, accuracy, ROI)
AWS GenAI infrastructure
Identify AWS services (for example, SageMaker JumpStart, Amazon Bedrock, Amazon Q)
Describe advantages of AWS GenAI services (for example, accessibility, speed, cost-effectiveness)
Describe benefits of AWS infrastructure (for example, security, compliance, safety)
Describe cost tradeoffs (for example, latency, availability, pricing models)
3. Applications of Foundation Models - 28%
Design considerations
Identify selection criteria for models (for example, cost, latency, multilingual, size)
Describe inference parameters (for example, temperature, output length)
Define RAG and its applications
Identify AWS services for embeddings and vector storage
Explain cost tradeoffs for customization approaches
Describe role of agents in multi-step tasks
Prompt engineering
Define prompt engineering concepts (for example, context, instruction, prompts)
Define techniques (for example, zero-shot, few-shot, chain-of-thought)
Identify benefits and best practices
Define risks (for example, prompt injection, jailbreaking)
Training and fine-tuning
Describe training stages (for example, pre-training, fine-tuning, distillation)
Define fine-tuning methods (for example, transfer learning, domain adaptation)
Prepare data for fine-tuning (for example, curation, labeling, RLHF)
Evaluate FM performance
Determine evaluation approaches (for example, benchmarks, human evaluation)
Identify metrics (for example, ROUGE, BLEU, BERTScore)
Evaluate business effectiveness
Evaluate applications using FMs (for example, RAG, agents)
4. Guidelines for Responsible AI - 14%
Responsible AI development
Identify features (for example, fairness, safety, robustness)
Use tools for responsible AI (for example, Guardrails)
Define responsible model selection practices
Identify legal risks (for example, bias, IP issues)
Identify dataset characteristics (for example, diversity, balance)
Describe bias and variance effects
Describe tools for monitoring bias (for example, SageMaker Clarify, A2I)
Transparency and explainability
Differentiate transparent vs non-transparent models
Describe tools (for example, Model Cards)
Identify tradeoffs between safety and transparency
Describe human-centered explainability principles
5. Security, Compliance, and Governance for AI Solutions - 14%
Secure AI systems
Identify AWS security services (for example, IAM, encryption, Macie)
Describe source citation and data lineage
Describe secure data engineering practices
Describe security and privacy considerations
Governance and compliance
Identify AWS governance tools (for example, AWS Config, CloudTrail)
Describe data governance strategies (for example, lifecycle, monitoring)
Describe governance processes and frameworks
