Chapter Zero · Beginner FAQ

What Is This "Knowledge Base" Every Company Is Building?

In meetings you keep hearing "we need to build a knowledge base" and "hook up RAG" — it sounds like a huge project. It's really just giving the AI a bookshelf it can look up anytime: the exam is open-book now, and it has to point you to the page it flipped to.

One-sentence answer

Slice company files into chunks and build an index. Before the AI answers, it finds the relevant passages, stuffs them into the conversation, then replies. It speaks from your docs, and it can cite the source for you.

Watch a demo · Three steps to get the knowledge base

The whole knowledge base runs in three steps. Click "Next" to move forward. Each step uses that bookshelf analogy from a moment ago.

📄Collect files: policy manuals, product docs, meeting notes — ordinary company files.
✂️Slice into chunks: long files become passage-sized "cards," one card per topic.
🗂️Put them on the shelf and index them: each card is registered with "what it's about," so you can look it up by meaning later.
Expense policy · passage 1Expense policy · passage 2Attendance rules · passage 1Product manual · passage 4Leave policy · passage 2Meeting notes · passage 7
🙋Someone asks: "What's our travel expense cap?"
🔍Check the shelf first: search by "similar meaning," and two cards hit (the ones that turn green below).
Expense policy · passage 1Expense policy · passage 2Attendance rules · passage 1Product manual · passage 4Leave policy · passage 2Meeting notes · passage 7
📌Note: this step is only "flipping the book" — the AI hasn't spoken yet.
📄Stuff the two passages into the conversation, join them with the question, and hand it to the AI.
💬The AI answers from the docs: the answer comes from your files, so it can guess less and make less up.
🔖Cite the source while it's at it: "Source: Travel-Expense-Policy.pdf" — open the original and you can check.
The same question · How much a knowledge base changes the answer

Ask the same line — "What's our travel expense cap?" — then switch the two buttons below and see where the AI's answers diverge.

Why not just feed it every file

Someone will ask: why go to all this trouble — just dump a few hundred company files on the AI. Two reasons:

🪑

The desk isn't big enough

There's a limit to how much the AI can "keep in view" at once — like a desk of fixed size. A few hundred files simply won't spread out. That desk is called the context window. This page explains why it "forgets".

💸

You pay by the word

Every word the AI reads is billed by token. Attaching every file to every question is like moving the whole library each time — your wallet gives out first. How token billing works is on this page.

So the knowledge-base idea is smart: leave the books on the shelf, and each time bring only the most relevant few pages into the exam. It saves money, and it fits.

The knowledge base's weak spot

A knowledge base lets the AI speak from the docs, but its ability to tell true from false is limited: if a file on the shelf is outdated, it answers from the old file; if two files contradict each other, it may copy half from each. What it guarantees is "there's a source you can check" — not "the source is trustworthy." So half the work of building a knowledge base is cleaning the files: delete what's outdated, merge what's contradictory. A clean shelf is a clean answer.

This page is the beginner version. To see how engineering actually slices chunks, builds the index, and raises the hit rate, the main course has a deeper lesson: Using RAG to ease hallucination.

✅ What this page wants to share with you