The knowledge base is the backbone of your agent’s configuration. It organizes topics and interactions that guide responses to user queries. Each topic typically represents a single FAQ or use case and contains three key parts:

  1. Topic name: A concise, descriptive title summarizing the topic.
  2. Content: The main information, including sample queries and model answers.
  3. Actions: Follow-up workflows triggered by the topic, like sending SMS messages or handing off to a live agent.

A well-structured knowledge base ensures your agent retrieves accurate information, performs tasks reliably, and delivers a consistent user experience.

​
Key components of a topic

​
Topic name

  • Clarity and precision: Use clear, descriptive titles such as “Opening Hours” or “Refund Policy” that reflect the topic’s content.
  • Analytics impact: Topic names influence search accuracy, dashboard analytics, and reporting.
  • Best practice: Avoid generic or overly broad names that can lead to misretrieval.

​
Sample questions

  • Sample questions help the model determine when to use this topic.
  • Include at least 5 phrased variations per topic to boost retrieval reliability and reduce fallback errors.

If a topic has too few or overly narrow examples, the model may skip it — even if the content matches. This often leads to fallback responses like “I don’t have information about that.”

Example (for “Refund Policy”):

  • “Can I get my money back?”
  • “How do I request a refund?”
  • “What’s your return policy?”

​
Content section

  • Core information: Add concise, actionable content that directly addresses user questions. Write in the tone you expect the agent to use.
  • Organize large topics: Break down complex topics into logical, smaller sections to ensure clarity and precision.

​
Actions section

The Actions section allows the agent to perform workflows beyond delivering answers.

  • Common actions:

  • Structured instructions: Clearly define what should happen when a topic is triggered.

    • Example: "If the user asks for emergency assistance, trigger a handoff to a live agent in the support queue."

Learn more: How to configure Actions

​
Best practices for building a knowledge base

​
Write clear, actionable content

  • Use direct language that’s easy for users to understand.
  • Keep content concise and avoid ambiguity.

​
Add diverse sample questions

  • Include 5–10 different phrasings of common user queries per topic.
  • These help clarify the scope of the topic for the model.

​
Simplify and structure topics

  • Group related questions under a single topic if they share similar context or answers.
  • Avoid overloading a single topic with too many unrelated intents.

​
Plan for risky or critical scenarios

  • Identify high-risk or regulated topics (e.g., refunds, complaints, emergencies).
  • Add specific instructions or actions to ensure they’re handled clearly and consistently.

​
Define and test actions

  • Use the Actions field to link workflows like SMS notifications, function calls, or handoffs.
  • Test actions thoroughly to ensure they behave as expected in real scenarios.

​
Global rules and tone

Use global rules to set overall tone and behavior across all topics.

Example:

  • Rule: “Respond in a friendly, concise tone. Avoid long explanations or unnecessary repetition.”

​
Iterate and improve

  • Monitor conversations and continuously update your KB based on real examples.
  • Add new sample questions to underperforming topics.
  • Refine unclear content or missed actions to improve response accuracy.

By following these best practices, you’ll create a knowledge base that’s accurate, reliable, and easy to scale.

Want to upload multiple topics at once? See our CSV import guide.