AWS AI Practitioner

Basic AI, ML, and deep learning concepts · Types of learning (supervised, unsupervised, reinforcement) · Common ML use cases and when to use AI · The ML development lifecycle · Data types and data preparation basics · Model evaluation metrics

Foundation models and large language models · Tokens, embeddings, and prompts · Generative AI use cases and limitations · AWS generative AI services (Amazon Bedrock, Amazon Q) · Advantages and challenges of generative AI · Business value and cost tradeoffs

Prompt engineering techniques · Retrieval Augmented Generation (RAG) and knowledge bases · Fine-tuning and model customization · Selecting and evaluating foundation models · Agents and orchestration · Vector databases and embeddings

Fairness, bias, and inclusiveness · Transparency and explainability · Responsible model selection and dataset choices · Tools for responsible AI (Guardrails, model evaluation) · Hallucinations and safety considerations · Human oversight

Securing AI systems and data · IAM and access control for AI services · Data privacy and protection · Governance and compliance frameworks · Monitoring and auditing AI workloads · Shared responsibility model for AI