# Knowledge Base
## Overview
### What is the Knowledge Base?
The Knowledge Base in Squarebox is a centralized system that equips your AI agent with the essential context, content, and communication style to provide meaningful and accurate responses to customer queries. It allows your team to manage resources effectively, personalize AI interactions, and continuously improve bot performance.
### Navigation
To access the Knowledge Base, go to:\
**Settings > Knowledge Base**
### Main Components of the Knowledge Base
The Knowledge Base is organized into **five key components** that work together to enhance AI capabilities:
#### 1. Business Resources
Upload and manage essential content such as guides, policies, FAQs, and manuals that the AI refers to when answering queries. The quality and breadth of these resources directly affect the AI’s accuracy.
#### 2. Tone Settings
Define the emotional and stylistic tone your AI should use, including predefined tones (Professional, Friendly, Direct, Empathetic) and custom instructions. This ensures the AI communicates in a way that matches your brand and context.
#### 3. Voice Settings
Customize the AI’s introduction or welcome message, making the agent feel more personable and aligned with your brand identity. A friendly or formal voice sets the right expectations for users.
#### 4. Unresolved Queries
Monitor and review questions your AI couldn’t answer effectively. This insight helps identify knowledge gaps and update resources to improve future responses.
#### 5. Agent Action Settings
Control when and how your AI agent interacts with users during chats, including auto-response rules, escalation to human agents, and chat behavior adjustments by topic.
***
### How These Components Work Together
By keeping these components up-to-date and well-configured, your Knowledge Base ensures that your AI agent delivers accurate, consistent, and helpful responses—boosting customer satisfaction and reducing the need for manual support.
## **Business Resources**
### What Are Business Resources?
Business Resources are essential documents, links, and guides that your AI agent refers to when answering customer queries. The more detailed and relevant these resources are, the smarter and more accurate your AI becomes.
***
### What You Can Do
* **Upload content** such as FAQs, policies, product manuals, or training materials.
* **Control access** by enabling or restricting the AI’s use of each resource.
* **Keep content fresh** by updating documents regularly with time-stamped revisions.
***
### How to Add a New Resource
1. Go to **Settings > Knowledge Base > Business Resources**.
2. Click the **“Add Resource”** button.
3. Enter a clear and descriptive **Name** for the resource.
4. Select the **Type**:
* **Link** — for online resources or external documentation
* **Document** — for PDFs or uploaded files
5. Toggle **Agent Access**:
* ✅ Enable this if you want the AI to refer to this resource.
6. Click **Save** to add the resource.
***
### Recommended Resource Types
* Product feature documentation
* Troubleshooting guides
* Employee onboarding manuals
* Company policies
* Training materials
***
### Pro Tips
* Use descriptive titles such as **“Refund Policy – India”** or **“Feature Walkthrough – Dynamic Campaigns”** for easy reference.
* Regularly audit and update your resources to maintain AI accuracy.
* Enable only the resources relevant to your AI’s scope to avoid confusion.
***
### Example View
| Resource Name | Type | Last Updated | Agent Access |
| ------------- | ---- | ---------------------- | ------------ |
| meta | Link | May 8, 2025 – 12:37 PM | ✅ Enabled |
## **Tone & Voice Settings**
### Tone Settings
Tone Settings allow you to define the emotional and stylistic tone your AI uses when communicating with users, ensuring responses fit your brand personality and context.
#### Predefined Tone Options
| Tone | Description | Use Case Example |
| ------------ | -------------------------------- | -------------------------------------- |
| Professional | Neutral, formal, informative | Policy replies, official communication |
| Friendly | Approachable and conversational | General product queries, onboarding |
| Direct | Short, focused, and efficient | Tech support, alerts |
| Empathetic | Supportive and emotionally aware | Customer complaints, delays |
***
#### Custom Instructions
Beyond tone presets, you can add custom instructions to guide how the AI structures responses.
**Examples:**
* “Always begin responses with ‘Hello \[Customer Name],’”
* “If the user asks for a guide, include a link and suggest next steps.”
***
#### Response Length Settings
You can also control how detailed your AI’s responses should be:
* **Concise:** Fast, brief answers (e.g., “Yes, that’s available.”)
* **Moderate:** Balanced detail (e.g., “Yes, that’s available. You can enable it under Settings > Preferences.”)
* **Detailed:** Comprehensive, multi-step replies.
***
#### Save Your Settings
Don’t forget to **click Save Settings** after making changes to ensure your tone preferences apply.
***
### Voice Settings (Agent Introduction)
This section controls how your AI introduces itself to users, which builds trust and aligns the assistant with your brand’s identity.
***
#### How to Customize Voice Settings
1. Navigate to **Voice Settings** inside the Knowledge Base.
2. Write a **Welcome Message** that your AI agent will use to start chats.
***
#### Example Welcome Message
> “Hi! I’m Ava, your AI assistant from Squarebox. How can I help you today?”
***
#### Why Voice Settings Matter
* Sets clear expectations for users
* Reinforces your brand tone (casual vs formal)
* Makes the AI feel more personable and helpful
### Improving AI with Unresolved Queries
### What Are Unresolved Queries?
Unresolved Queries are customer questions or requests that the AI agent couldn’t answer properly or didn’t understand. Tracking these helps identify knowledge gaps or missing resources.
***
### How to Use Unresolved Queries
* **Regularly review** unresolved queries to spot recurring questions or issues.
* **Identify gaps** where your AI lacks information or resources.
* **Add or update resources** in the Knowledge Base to cover these gaps.
* **Export unresolved queries** to share with your team for training or analysis.
***
## Customizing AI Agent Actions
### What Are AI Agent Actions?
AI Agent Actions allow you to control how and when your AI engages with users during chat conversations. This customization helps balance automation with human support for the best customer experience.
***
### Example Rule
| Query | Description | Last Updated | Category | Status |
| ------------------------ | ------------------------- | ------------ | -------- | --------- |
| Respond to chat all time | AI agent is always active | Apr 29, 2025 | All | ✅ Enabled |
***
### Key Settings
* **Auto-Response Conditions:** Define when the AI should reply automatically and when it should wait.
* **Escalation Logic:** Set rules to route conversations to human agents, for example, after 3 failed AI replies.
* **Chat Behavior by Category:** Adjust AI tone or urgency based on the topic (e.g., billing questions vs feedback).
***
### Important Considerations
⚠️ Use these settings carefully to maintain a good balance between automation and human intervention, ensuring optimal customer satisfaction.
***
### Best Practices for Admins
* Keep resources updated regularly to improve AI accuracy.
* Use Friendly or Empathetic tones especially in support or sensitive scenarios.
* Review unresolved queries weekly to close knowledge gaps.
* Personalize voice settings for a more human-like AI experience.
* Enable only relevant resources to keep AI responses focused and clear.
***
### Impact of a Well-Maintained Knowledge Base
| Benefit | Result |
| ------------------------------ | ----------------------------------------- |
| AI gives more accurate replies | Increased customer satisfaction |
| Fewer unresolved queries | Less need for human intervention |
| Brand-consistent communication | Professional, unified customer experience |
| Faster query resolutions | Boosted productivity & trust |
***
{% embed url="" %}
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://squarebox-ai.gitbook.io/squarebox/knowledge-base.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.