Contents
Technical writers have always written for humans. But in 2026, documentation is no longer consumed only by people. It is consumed by AI systems; RAG pipelines, chatbot knowledge bases, and AI-powered Search assistants now consume the same content your users do; and they read very differently.
A well-crafted narrative that guides a human reader through a complex installation workflow may be completely invisible to a retrieval system scanning for discrete, verifiable facts. Metadata that human readers skip over is what AI depends on.
Drawing on a career spanning software development, QA, business analysis, and technical writing, this session explores writing for dual audiences; human and machine. It offers a practical framework for structural choices that serve both, identifies where their needs genuinely conflict, and provides a checklist for auditing existing documentation for AI readiness, without triggering a full content rewrite.
Takeaways
How AI systems read documentation differently from humans, which structural choices serve both audiences, and a practical checklist to audit existing content for AI readiness so AI Search shows accurate outcomes.
Prior knowledge
Attendees should have at least 2–3 years of active technical writing experience and a basic familiarity with how AI tools are currently used in documentation workflows; for example, using AI assistants for drafting, editing, or content search. No engineering or machine learning background is required. The session does not assume knowledge of RAG systems or LLMs; all technical concepts are explained in documentation terms.