Documentation/Knowledge Base/How the AI Uses Knowledge

How the AI Uses Knowledge

What you'll learn: How your knowledge base content powers AI responses across your practice.

The Connection

Your knowledge base directly feeds both:

  • Voice AI — The phone assistant
  • Chat AI — The in-app assistant

When either AI needs to answer a question, it searches your knowledge base.

How Questions Are Answered

Step 1: Understanding the Question

The AI interprets what's being asked:

  • Recognizes key words
  • Understands intent
  • Matches to known topics

Step 2: Searching Knowledge

The AI looks for relevant content:

  • FAQs that match
  • Related information
  • Applicable answers

Step 3: Crafting a Response

The AI creates an appropriate answer:

  • Uses your content as a base
  • Adapts to the conversation context
  • Maintains a natural flow

Step 4: Delivering the Answer

The response is provided:

  • Over the phone for Voice AI
  • In text for Chat AI
  • With appropriate tone

Why Your Content Matters

The quality of AI responses depends on your content:

  • Good content → Accurate, helpful answers
  • Missing content → AI can't answer
  • Wrong content → AI gives wrong answers
  • Outdated content → AI gives old information

What Happens When Information Is Missing

If the AI can't find an answer:

  • It acknowledges it doesn't know
  • It offers alternatives (take a message, transfer)
  • It doesn't make up information

This is why filling gaps in your knowledge base matters.

Prioritization

If multiple pieces of content might apply:

  • More specific matches take priority
  • Recent updates may be preferred
  • The AI chooses the most relevant

Continuous Improvement

The AI gets better as you:

  • Add more FAQs
  • Update outdated content
  • Fill gaps identified from calls
  • Refine answers based on feedback

Testing Your Knowledge Base

To see how the AI uses your content:

  • Ask the Chat AI questions you've added
  • Listen to call recordings
  • Check if answers match your content
  • Note any discrepancies

The Feedback Loop

  1. You add content to knowledge base
  2. AI uses it to answer questions
  3. You review how questions were handled
  4. You improve content based on real interactions
  5. AI responses get better

Next: Settings Overview | Previous: Managing Knowledge Content

← PreviousManaging Knowledge Content