What This Section Is For
This section provides practical guidance for documenting AI-powered products and features.
Each guide focuses on how to think about documentation for a specific type of AI system. These are not templates or step-by-step instructions. They are meant to help you decide what to explain, what to clarify, and what to make visible.
The guides assume you already understand:
- what AI is and isn’t
- the AI category you’re working with
- basic technical writing principles
Who These Guides Are For
These guides are for:
- Technical writers responsible for documenting AI-powered features
- Writers supporting product, UX, or engineering teams building AI functionality
- Documentation professionals navigating new AI-related responsibilities
They are written for people shaping user understanding, not designing models.
How to Use These Guides
If you are unsure how to classify your product or feature, begin with the AI Product Landscape.
Then:
- Read the AI category article that matches what you’re documenting.
- Use the corresponding Documenting X guide to shape:
- onboarding content
- explanations of behavior and limitations
- trust and expectation setting
You don’t need to read every guide. Choose the one that aligns with the system you are working on.
What These Guides Cover
The guides address:
- Documentation challenges unique to different AI product categories
- Common user misunderstandings and failure points
- How AI behavior affects onboarding, trust, and long-term use
- Where traditional documentation patterns break down
They focus on writing decisions, not feature marketing.
What These Guides don’t Cover
The guides don’t cover:
- AI implementation details
- Model architectures or training methods
- API reference documentation
- General technical writing best practices
For foundational writing principles, see Technical Writing Best Practices.
The Documentation Guides
Each guide corresponds to an article AI product category or system behavior pattern.
Documenting AI Assistants, Copilots, and Agents
Guidance on documenting task-oriented, embedded AI features, with attention to onboarding, context, trust, and autonomy.
Documenting Generative AI
Guidance on documenting AI features that generate content, focusing on variability, examples, and expectation setting.
Documenting Conversational AI
Guidance on documenting conversation-first systems, including dialog behavior, misunderstanding, and recovery.
Documenting Predictive AI
Guidance on documenting predictive systems, with attention to uncertainty, confidence, and interpretation of results.
Documenting AI Search and Retrieval
Guidance on documenting AI-powered search systems, including scope, ranking, and “why didn’t it find this?” questions.
Documenting AI Workflow Automation
Guidance on documenting AI-driven automation, including triggers, actions, and recovery paths.
Documenting AI Personalization Systems
Guidance on documenting adaptive and personalized AI behavior, with focus on transparency and user control.
Documenting AI Safety and Guardrails
Guidance on documenting constraints, refusals, and limits in a way that builds trust and reduces confusion.
Documenting Agentic AI Systems
Guidance on documenting AI systems that can plan and execute multi-step tasks by calling tools and operating within defined constraints.
How This Section Fits Into the Series
This section provides applied writing guidance.
It builds on the conceptual foundation in What Artificial Intelligence Is (and Isn’t) and the classification framework in the AI Product Landscape.
It complements, but does not replace, Technical Writing Best Practices.
Where to Go Next
- Choose the guide that matches the AI feature you’re documenting.
- Revisit Technical Writing Best Practices if you need foundational writing support.
- Use the Getting Started page if you are unsure where to begin.
Documenting AI is less about technical precision and more about shaping understanding over time. Use these guides as thinking tools, not rules.