Knowledge base and RAG: answers from approved facts only

Intermediate⏱ 18 min
🔒Join free to watch this video lessonFree account · takes 30 seconds · no card neededJoin free & watchAlready a member? Log in
Give the agent a searchable knowledge base, force it to ground answers in it, and test for hallucination.

In this lesson you’ll learn to

  • I wrote Q&A chunks in spoken style
  • I added the grounding instruction
  • I ran answerable, plausible and trap questions

Outcome of this lesson

You can add a knowledge base so the agent answers from your approved information — and says “I’m not sure” when the answer is not there.

Why not paste everything into the prompt?

• Long prompts slow every turn and distract the model.

• Information changes — prices, hours, policies — and prompts become stale.

• The agent may mix facts or answer confidently from general training instead of your policy.

RAG in one paragraph

Retrieval-Augmented Generation: your documents are split into chunks and indexed. When the caller asks something, the system searches for the most relevant chunks and hands them to the model, which answers using only those chunks. Most voice platforms offer a built-in knowledge base or a “search” tool; you can also expose your own search endpoint as a tool.

Prepare

🔓 Unlock the full lesson — free

Create a free account to read the step-by-step guide, see the worked example, take the quiz, download templates and track your progress.

💬 Questions about this lesson?
Ask the community — practitioners and the BFOUNT team answer.
Ask a question
Was this lesson helpful?