USE CASE — KNOWLEDGE BASE

A Help Center That Actually Answers Questions

Install the Documentation template, structure your articles, and let semantic search and an AI chatbot do what a support search box usually fails at.

Start Building Free

Semantic search, not keyword matching

AI chatbot answers with cited sources

Built-in Documentation template

THE CORE PROBLEM

Your Terminology Isn't Their Terminology

A visitor searching "why won't my export finish" and an article titled "Handling Large Data Exports" share almost no words. Keyword search misses this constantly. Semantic search matches on what the query means, not what it says.

Matches Concepts

Matches concepts, not shared keywords

Falls Back Automatically

Falls back to keyword search if needed, and reports which mode answered

Query Logs Show Gaps

Query logs show exactly which questions your docs fail to answer

An Answer, Not Just a List of Links

Zunoy's AI chatbot answers a visitor's question directly from your published articles, with the source articles cited alongside the answer. It cannot answer from anything you haven't published.

See Semantic Search

Answers generated from your actual published content

Every answer carries the sources it drew from

Weak-confidence answers are visible, not hidden

Common Questions

Does Zunoy include a documentation content structure?

Yes. The built-in Documentation template installs a Doc Article content type in one step, giving you a real starting structure rather than an empty content type to model from scratch.

Can the chatbot invent an answer that isn't in my docs?

No. Answers are generated only from your published articles, with sources cited alongside the answer, so the chatbot cannot surface information that was never published.

Will search actually find the right article?

In most cases, yes, because search matches on meaning rather than requiring the visitor's exact wording to overlap with the article's title or body — which is the usual failure point for support search.

How do I know what to write next?

Query logs record what visitors search for, including questions that return weak matches, which functions as a prioritized list of documentation gaps drawn from real demand.