A chatbot knowledge base is the set of information the bot draws on when it answers. For a business website that usually means service pages, pricing, FAQs, opening hours, policies, and contact details. The language model provides the ability to write, and the knowledge base provides the facts. Without one, even a capable large language model (LLM) can only answer from general training data, which knows nothing about your business.
What Belongs in a Knowledge Base
The strongest knowledge bases are built from content customers already ask about. A quick test is to read your last month of emails and phone notes, list the questions that came up more than once, and check that each one has a clear answer somewhere on your site.
- Services and prices: what you offer, what it costs, and what is included.
- Logistics: hours, service area, parking, booking steps, and turnaround times.
- Policies: cancellations, refunds, warranties, and payment methods.
- Contact routes: phone, email, and booking links for questions the bot should hand over.
How a Chatbot Uses It
When a visitor asks something, the chatbot searches the knowledge base for the passages most related to the question and passes them to the model along with the message. The model then writes the reply from that material. This is far more flexible than the decision trees described in rule-based vs AI chatbots, because nobody has to predict every question in advance. It does mean answers are only as good as the pages behind them. Outdated prices on a page become outdated prices in the chat.
Building One in AI Chatbot
In AI Chatbot, the knowledge base comes from your WordPress site. You choose which pages to include, add custom URLs, and set priority content. When you edit those pages, the Re-index button refreshes what the bot knows. Each option is listed on the AI Chatbot features page, and how AI Chatbot works shows the setup in order.
