What a knowledge base is and why AI needs one
A chatbot is only as good as the documents it reads. If your information is scattered across emails, old PDFs and a colleague's head, the AI will answer incompletely or contradict itself.
Building the knowledge base is often the most useful part of an AI project, for your people too: it forces you to write down, once and clearly, things that used to be explained out loud every time. It needs to be kept up to date, and someone has to be responsible for it.
An example in an SMB
A company that makes management software puts the answers to its most frequent tickets into a single knowledge base, written by the support team. The chatbot answers from there, and when it can't find the answer it opens a ticket. Every month the support team adds answers for the new cases.
The most common mistakes
- Uploading everything without choosing: outdated documents confuse the assistant.
- Not saying who updates it: after a few months it is out of date.
- Writing for insiders: answers must be understandable to the customer.
- Keeping different versions of the same document in different places: the assistant doesn't know which one is valid.
Why it matters to you
A well-made knowledge base is useful even without AI: new hires learn faster, colleagues stop asking each other the same things, customers find answers on the website. The assistant becomes the most convenient way to consult it, not the only reason to have it. If the AI project stops, the work you have done is still useful.
All the words of AI explained simply are in the glossary.
Frequently asked questions
Where do I start?
With the questions customers ask most often. Write a clear answer for each one, then expand.
What format should I use?
Plain text, well divided by topic, works better than long, formatted PDFs.
How often should it be updated?
Whenever something changes that customers ask about: prices, terms, opening hours, products. Every so often, review the questions the assistant couldn't answer: they show what is missing.
Read next
What is RAG: AI that answers from your documents
RAG (Retrieval Augmented Generation) is a technique where, before answering, the AI searches your documents for the useful passages and uses them as the basis for its answer.
GlossaryWhat an AI chatbot is and how it differs from the old ones
A chatbot is a program that holds written conversations with people. An AI chatbot uses a language model to understand freely worded questions and answer in a natural way.
GlossaryWhat AI hallucinations are and how to reduce them
An AI model hallucinates when it produces false or invented information, such as a date, a source or a quotation, and presents it with the same confidence as a true one.
GlossaryWhat a vector database is and when you really need one
A vector database is a store built to hold embeddings and quickly find the ones closest to a question, that is, the texts most similar in meaning.
Want to understand what your business really needs, without the jargon? Let's talk.
Text and graphics created with the support of AI. Method and content are AIR's.