Knowledge base software that your help desk and AI agent both use
Ticketagent gives you a public help center for customers, internal articles for your team and one source of truth for the AI agent that drafts and resolves tickets. Every answer the AI gives points back to an article you wrote and approved, so improving the knowledge base improves every reply.
A knowledge base is the foundation of good support
Most support teams answer the same twenty questions over and over. A knowledge base turns those answers into articles that customers can find on their own, that new agents can learn from in their first week, and that the AI agent can quote with confidence.
The problem with most setups is that the help center lives in one tool, internal notes live in a wiki, and the help desk knows about neither. Agents copy and paste between tabs, articles drift out of date and nobody knows which pages actually help. Ticketagent keeps the knowledge base inside the help desk, next to the tickets it is meant to prevent.
- One editor for public and internal articles
- Articles linked to the tickets they answered
- The same content feeds your help center, your agents and the AI agent
- Feedback on every article, so you know what to fix first
A public help center your customers can actually search
Publish articles to a branded help center on your own domain. Customers search by the words they use, browse by category and land on a clear answer instead of opening a ticket. The web form widget shows matching articles while the customer types, so a question about resetting a password gets the reset article before it becomes a ticket.
- Categories and sections that match how your product is organized
- Search across titles, body text and tags
- Article suggestions inside the web form and live chat widgets
- Your logo, colors and domain on every page
- On Business, up to 5 brands, each with its own help center
Internal articles for the answers customers should not see
Not every answer belongs in public. Refund exceptions, escalation contacts, known bugs with a workaround, steps that need admin access: these belong in internal articles that only your team can read. Agents find them from the ticket sidebar, and the AI agent can use them when drafting a reply for a human to review, without ever showing the internal text to the customer.
- Visibility per article: public or internal
- Internal articles appear in the agent sidebar next to the ticket
- Permissions control who can edit and who can publish
The knowledge base is the source the AI agent cites
The AI customer service agent in Ticketagent does not invent answers. It drafts replies grounded in your articles, and the agent reviewing the draft sees exactly which article the answer came from. If the article is right, the reply is right. If the article is missing or outdated, the AI says so and hands the ticket to a person with a short summary.
This changes how you think about writing articles. An article is no longer just a page for customers who search. It is the instruction set for every AI draft and every automatic resolution on that topic. When you fix a step in an article, the next reply already uses the corrected step.
- Every AI draft shows the cited article to the agent
- Auto-resolution only for intents you approve, above a confidence threshold you set
- When no article fits, the ticket goes to a human with a summary
- Customer data is not used to train shared models
Article suggestions from resolved tickets
Your best articles are already written, scattered across resolved tickets. Ticketagent looks at tickets your team closed and suggests new articles where the same question keeps coming back and no article covers it. Each suggestion comes with the questions customers asked and the reply that solved them, so the first draft takes minutes, not an afternoon.
- Suggested topics ranked by how often the question comes up in your tickets
- A draft article built from the reply that resolved the ticket
- An editor reviews, edits and publishes; nothing goes live on its own
- Links from the article back to the tickets it came from
Article feedback that tells you what to fix
Each article asks readers whether it helped. A "no" can include a short comment. Inside Ticketagent you see which articles get the most negative feedback, which articles are followed by a ticket on the same topic anyway, and which ones the AI agent cites most often. That gives you a short, ordered list of articles worth rewriting instead of a vague sense that the help center needs work.
- Helpful and not helpful votes with optional comments
- Articles most often followed by a new ticket on the same topic
- Articles most cited by the AI agent in drafts and resolutions
- Reports next to your ticket volume and AI resolution rate
Article limits per plan
The knowledge base is part of every plan. What changes between plans is how many articles you publish and whether internal articles are included.
| Plan | Public articles | Internal articles |
|---|---|---|
| Starter | Up to 100 | Not included |
| Team | Unlimited | Unlimited |
| Business | Unlimited, up to 5 help centers | Unlimited |
| Enterprise | Unlimited | Unlimited |
A hundred public articles covers the core questions of most small teams. When you need internal articles for your agents or a larger public library, Team removes the limit.
How to set up your knowledge base in Ticketagent
- Step 1: List the 20 questions your team answers most often. Your recent tickets are the fastest source.
- Step 2: Create categories that match your product areas, not your org chart.
- Step 3: Write or import one article per question, with the answer in the first two lines.
- Step 4: Mark articles as public or internal and publish the help center on your domain.
- Step 5: Turn on AI reply drafts and check which articles the AI cites in its first week of drafts.
- Step 6: Review article suggestions from resolved tickets every week and fill the gaps.
Our guide on how to build a knowledge base walks through article structure, naming and review cadence in more detail.
Knowledge base software questions
Yes, you can paste existing articles into the editor, or on Team and up import them through the REST API.
It uses internal articles for drafts that an agent reviews, and it never shows internal text to a customer.
The AI agent does not guess; it hands the ticket to a human with a summary and the gap shows up in article suggestions.
Yes, every plan publishes the help center on your own domain with your logo and colors.
Starter includes up to 100 public articles, and Team, Business and Enterprise include unlimited public and internal articles.
Knowledge base included in every plan
Starter is $29 per agent per month ($24 billed yearly), Team is $59 ($49 billed yearly), Business is $99 ($82 billed yearly), and Enterprise has volume pricing through sales. The knowledge base, the help center and AI reply drafts are part of the seat price on every plan. See help desk software pricing for the full comparison.