AI knowledge base

How knowledge facts work, Teach AI, web imports, per-package scoping, duplicate detection, and source revocation.

The knowledge base is what makes the AI assistant accurate. It is a collection of facts: short, verifiable statements about your products that the AI injects into its answers.

Facts

Each fact has:

  • Content: up to 2,000 characters
  • Type: feature, config, command, requirement, troubleshooting, or other
  • Scope: global (applies to every ticket) or linked to one Tebex package
  • Source: where it came from (manual entry, a ticket conversation, or an imported URL)
  • Confidence: how certain the AI was when it extracted the fact

There is no cap on the number of facts, on any plan.

Permissions

PermissionAllows
ai.viewBrowse facts, see which facts an AI reply used
ai.teachExtract facts from ticket conversations (Teach AI)
ai.knowledge.manageAdd, delete, import, and revoke facts

How the AI picks facts

When generating a reply, the AI uses every global fact plus, if the ticket is linked to a Tebex package, the facts scoped to that package. Scoping facts to packages keeps answers precise: a fact about script A is never used to answer a ticket about script B.

Every AI reply records which facts it used. Staff can open the facts used popover on any AI message to see them, along with the AI's confidence score.

Teach AI

Teach AI extracts facts from a ticket conversation:

  • The conversation needs at least 2 human messages. Internal notes, AI-generated messages, and system messages do not count.
  • The last 20 human messages are analyzed.
  • The AI proposes a list of facts; nothing is saved automatically. You review, edit, or discard each fact before saving.

Facts saved from a ticket keep a reference to it. If the ticket is later deleted, the facts survive.

Adding facts manually

Staff with ai.knowledge.manage can write facts directly. When you submit one, Joely compares it semantically against existing facts: if a near-identical fact exists (around 90% similarity), you get a warning showing the match. The warning is not blocking, you can save anyway.

Web import

Import knowledge from any public web page:

  1. Submit a URL (one per import). Only public HTTP/HTTPS addresses are accepted; private and internal hosts are blocked.
  2. Optionally provide CSS selectors to target specific page sections or remove noise (up to 10 selectors each).
  3. A background job fetches the page (content capped at 100,000 characters), extracts facts with AI, and flags semantic duplicates.
  4. The extracted facts arrive unverified: you review each one (approve, edit, or discard) before they become part of the knowledge base.

Job progress is streamed live to the dashboard. Failed jobs appear in the jobs view and can be retried; completed and failed jobs can be deleted without affecting facts that were already saved.

Provenance and source revocation

Because every fact tracks its source, you can clean up in bulk when a source turns out to be unreliable:

  • Revoking a ticket source deletes every fact extracted from that conversation.
  • Revoking a web source deletes every fact imported from that URL, across all import jobs.

Joely shows you how many facts will be removed before you confirm. Revocation is permanent.

The documentation suggestion

If staff reject 3 AI drafts in a row (see approval mode), Joely surfaces a one-time suggestion to import documentation, since repeated rejections usually mean the AI lacks knowledge rather than judgment. Approving any draft resets the counter; each staff member can dismiss the suggestion permanently.

Cookies & Privacy

We use cookies to make your experience on this website better.