A course I built in Articulate Rise and delivered live at a work conference — making the case that structured content isn't just tidy housekeeping. Done right, your metadata is what lets AI actually find and use your content. And I didn't just argue it; I proved it.
Why DITA, why now
DITA (the Darwin Information Typing Architecture) is a standard for structuring and tagging content so it's consistent, reusable, and machine-readable. For years that mattered most for documentation reuse. But in the AI era it matters for a new reason: AI search engines are only as good as the metadata they're fed.
The course's core message was exactly that — getting your metadata right is what makes AI-powered search fast and accurate. Garbage in, garbage out; structure in, answers out.
The experiment that proved it
Rather than just assert the benefit, I ran a real test. We took our entire course curriculum — about 300 assets — and made them DITA compliant to load into our AI search engine, AIESA. Then we compared the engine's performance with and without the DITA-compliant documents:
Faster, more accurate, and better at retrieving the right knowledge — across all three measures. The structure paid off measurably, not just in theory.
Key takeaways
- Metadata is AI infrastructure. Structured, well-tagged content is what makes AI search work.
- Compliance is worth the effort — a 300-asset conversion measurably improved speed, accuracy and retrieval.
- Prove it, don't just preach it. An A/B test turned "DITA is good practice" into "DITA wins, here's the data."
- Structured content and AI readiness are the same investment, not competing priorities.
Built in Articulate Rise · Delivered live at a work conference.