Different AI surfaces. One governed source.
Public chatbots, employee copilots, help centers, docs assistants, and sales bots fail in different ways when they share a pile of unreviewed documents. These pages cover the live Wiki workflows for each surface.
AI knowledge use cases
AI knowledge use cases for support chatbots, internal copilots, help centers, docs assistants, and sales assistants that share one verified source of truth.
A support chatbot should only say what you approved.
Public chatbot answers come only from verified public-safe collections. The bot refuses when there is no approved source. Citations go out through Wiki Connect or the Answer Playground.
Employee copilots should not see everything.
Employee copilots stay scoped to internal collections. Public-safe knowledge and internal notes stay split. Each AI Application gets its own policy.
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.
Your docs assistant should quote the published page.
Docs assistants answer from verified public docs collections, cite the page, and refuse when the help center does not cover the question.
Sales AI should only repeat approved claims.
Sales assistants retrieve verified commercial knowledge, refuse prohibited claims, and cite the approved source when they answer.
Solutions for the problems that show up after launch.
Stop the chatbot from making things up.
Wiki cannot guarantee no hallucinations. It requires source-backed answers, refuses when nothing is approved, and keeps unverified and internal content out of public bots.
Do not let the chatbot invent a refund.
Stop a chatbot invented refund. Import the policy URL, verify it, put it in a public-safe collection, attach an AI Application policy, test in the Playground, then serve a cited Wiki Connect answer or refuse.
Retrieval is not an AI source of truth.
An AI source of truth is verified, permissioned knowledge with a policy per application. Wiki Management for AI Applications is that layer, not a second copy of your docs.
If it cannot cite a source, it should not answer.
Require AI answers to cite approved sources or refuse. Wiki tests citation behavior in the Answer Playground and serves it through Wiki Connect.
RAG needs a governance layer.
RAG retrieves. Wiki governs whether retrieved documents are verified, allowed, and safe to cite for each AI application.
Stale docs are an AI outage waiting to happen.
Stale docs become wrong AI answers at scale. Wiki tracks review cadence, source changes, and knowledge health before AI reuses a page.
Helpdesk AI needs governed knowledge behind it.
Ground Intercom and Zendesk bots in verified Wiki knowledge via Connect adapters and private apps. Cite or refuse instead of inventing refunds.
Your AI is already talking.
Make sure it knows what it is talking about.