What is a knowledge base?
A knowledge base is the published collection of articles a customer support team writes so customers can answer their own questions before contacting the team.
Updated June 19, 2026Also available in: Deutsch, Português
A knowledge base is the published collection of articles a support team writes so customers can answer their own questions before contacting the team. The articles are public, searchable, and organized by category. Customers find them through a portal, through search engines, or through the AI bot when it cites a relevant article in a chat.
Why a knowledge base matters
Most customer questions repeat. The same 50 questions account for 80 percent of inbound for most support teams. If the answer to those questions is published, customers can find it themselves and the team handles only the harder cases.
A good knowledge base reduces ticket volume, speeds up agent responses (because the agent can link to an article), and feeds the AI bot with grounded answers.
What a modern knowledge base includes
- Article editor: a clean editor with formatting, code blocks, embedded videos, and sometimes AI-assisted drafting.
- Categories and search: articles organized by topic with full-text search.
- Multi language: the same article in multiple locales, with hreflang tags for SEO.
- Versioning: draft, published, archived states with edit history.
- Custom domain: the help center runs on your own subdomain (help.example.com) with your branding.
- Analytics: view counts, search terms with no results, articles that lead to ticket creation anyway.
- AI grounding: articles feed the AI bot's context so the bot answers from the same source the customer would read.
The relationship between knowledge base and AI
The two are often sold separately, but they should not be. If your AI bot is grounded in your knowledge base, every article you write makes the bot smarter. The team writes one article, and that article serves both human readers and the AI.
The flip side is that articles missing from the knowledge base are also missing from the AI. If a question keeps coming up in chat and the bot can't answer it, the fix is to write the article. The bot reads it on the next conversation.
What "self serve" looks like in practice
A customer hits a problem at midnight. They go to your website. They type their question into the chat widget. The bot searches the knowledge base, finds the closest article, and replies with the answer plus a link.
If the bot's answer matches, the customer is unblocked without an agent ever touching the conversation. If not, the bot hands off to a human, who picks up in the morning with the full chat history and the article links the bot already tried.
How to evaluate a knowledge base feature
Three questions matter most:
Does it actually feed the AI? Some vendors sell the knowledge base separately from the AI. That sometimes means the bot has to be retrained when articles change, or that articles feed the bot only after a manual sync. Check whether grounding is automatic.
Can it run on a custom domain? SEO matters. If the knowledge base lives on the vendor's domain, you can't build link equity to your own brand.
Is migration possible? If you already have articles on Zendesk, Freshdesk, Intercom, or Crisp, the new platform should import them in one click with redirects so SEO transfers.
Where Deskwoot fits
Deskwoot's knowledge base is one product with the AI bot. Every published article is automatic grounding for the bot, no manual sync. Custom domains, multi language (English, German, Portuguese, Hindi), and one click migration from major competitors are included in every paid plan.