One verified source for the help center and the chatbot.
Publish public docs from the same verified sources the chatbot may use. Customers read the page. The bot cites the same page, or it refuses.
Publish public docs from the same verified sources your support chatbot may use. One review workflow, two surfaces, no invented help-center answers.
Two sources of truth is how help centers and chatbots diverge.
Most teams publish a help center in one tool and feed the chatbot a different export, a Notion dump, or last month's PDF. The page says 30 days. The bot says 60. Support spends the afternoon explaining which one is real.
A help center and a chatbot should share a source of truth. That does not mean pasting the same markdown into two products. It means one verified source, reviewed once, then published as docs and allowed for the public chatbot.
When the policy changes, both surfaces should move together. If only the help center updates, the chatbot keeps citing the old file. If only the bot's index updates, the public page is stale.
What to do to keep docs and the chatbot aligned
Publish from the same verified collections the chatbot is allowed to retrieve.
- Step 1
Import or write the canonical page once
URL import works when the policy already lives on a public URL. Otherwise create the document in Wiki and treat that as the source you will verify.
- Step 2
Verify it for AI use
Verification is the gate. Unreviewed drafts do not become public docs and do not become chatbot answers.
- Step 3
Publish public docs from that source
Wiki publishes public docs from approved knowledge. The help center is not a separate CMS with a second copy of the refund page.
- Step 4
Allow the same collection on the chatbot policy
The public chatbot retrieves from the public-safe collection that backs the docs. It does not get a parallel unreviewed index.
- Step 5
Test both surfaces against the same questions
Open the published page. Ask the same question in the Answer Playground. The citation should be that page, or the bot should refuse.
How Wiki publishes docs and chatbot answers from one source
Public docs are live. A docs assistant that sits on the published site is planned, not shipped. The chatbot path today is Wiki Connect or the Playground.
Verified sources
The same review workflow prepares knowledge for docs and for AI retrieval.
Public docs
Publish approved knowledge as a public help center from the collections you already govern.
Public-safe collections
The chatbot's AI Application policy points at the same public-safe collections that feed the docs site.
Answer Playground
Confirm the bot cites the published page before you expose Wiki Connect to production traffic.
Knowledge Health
See whether the workspace is actually ready to publish and retrieve, not just full of files.
Questions teams ask first.
- Can the help center and the chatbot use the same source of truth?
- Yes. Verify the source, keep it in a public-safe collection, publish public docs from that collection, and attach the same collection to the chatbot's AI Application policy.
- Does Wiki include a docs assistant on the help center?
- Public docs publishing is live. A docs assistant on the published site is planned and not claimed as available today. Test answers in the Answer Playground and serve them through Wiki Connect.
- What if the help center and the bot disagree?
- That usually means they do not share a verified source. Put the canonical page in Wiki, publish docs from it, and point the chatbot policy at that collection. Then re-test the question in the Playground.
- Can I import an existing help-center URL?
- Yes. URL import brings the public page in as a source. Verify it, place it in a public-safe collection, then publish and retrieve from that collection.
Essays, tools, and the product.
From the blog
What Is Wiki Management for AI Applications?
Wiki Management for AI Applications governs what AI can use, who can see it, whether it is current, and whether an answer can be trusted.
The Complete Guide to Citation-Backed AI Support
Customer-facing AI should show where the answer came from, know when it lacks an approved source, and escalate when documentation cannot support a reliable answer.
RAG Is Not a Source of Truth
Retrieval can find context. It does not decide whether that context is current, approved, allowed for the audience, or safe to cite.
Tools
Your AI is already talking.
Make sure it knows what it is talking about.