Loading demo links...
What to study for each D-PEN-F-A-00 domain
AI and NLP Fundamentals
This domain covers how large language models and natural language processing work at a foundational level, including tokens, context windows, embeddings, and sampling controls such as temperature. It also covers core NLP tasks and limitations such as hallucination that prompt engineers must understand.
Key skills
LLM concepts, tokens and tokenization, context windows, embeddings, training versus inference, temperature and sampling, hallucination awareness, core NLP tasks
Legal and Compliance Requirements
This domain covers the legal and compliance duties that apply when using generative AI at work, including data privacy, protection of confidential information, and intellectual property risks in AI-generated content. It also covers bias, accountability for published output, and transparency obligations.
Key skills
Data privacy and personal data handling, confidential information protection, intellectual property and copyright risk, bias and discrimination awareness, disclosure and human accountability
Introduction to prompt structure and formatting
This domain covers the building blocks of an effective prompt, including instructions, context, input data, delimiters, and output indicators. It also covers system and user prompts, specifying format, length, audience, tone, and language, and using templates and placeholders for reuse.
Key skills
Instruction writing, context provision, delimiters and sectioning, system versus user prompts, output format specification, length and tone constraints, grounding instructions, prompt templates and placeholders
Prompt Engineering Techniques
This domain covers practical techniques for improving model results, including zero-shot and few-shot prompting, chain-of-thought reasoning, role prompting, and prompt chaining. It also covers decomposition, self-consistency, hallucination reduction, iterative refinement, and evaluating prompt quality.
Key skills
Zero-shot prompting, few-shot examples, chain-of-thought, role prompting, prompt chaining, task decomposition, self-consistency, iterative refinement, prompt evaluation
Professional Practices for Prompt Engineers
This domain covers how prompt engineering is practiced professionally, including documenting and versioning prompts, building test sets, and measuring quality with appropriate metrics. It also covers production monitoring, collaboration with subject-matter experts, responsible AI use, and managing shared prompt libraries.
Key skills
Prompt documentation and version control, test set design, evaluation metrics, output verification, production monitoring and logging, expert collaboration, responsible AI practice, prompt library management