Managing Agent Knowledgebases

What are Knowledge Bases?

Knowledge bases are content sources that AI agents use to provide accurate, context-aware responses. CampusMindAI agents can connect to multiple knowledge sources including web pages, uploaded files, and cloud storage services like Google Drive. All content is automatically ingested into Azure AI Search for efficient indexing and retrieval.

Why Manage Knowledge Bases?

• Content Control: Monitor what information your agents can access

• Quality Assurance: Track synchronization status and identify issues

• Performance Monitoring: View content size and processing metrics

• Lifecycle Management: Update, resync, or remove outdated content

• Multi-Source Integration: Centralize content from web, files, and cloud storage

Knowledge Source Types

CampusMindAI supports three primary knowledge base types:

1. Web Pages: Crawled website content using configurable strategies

2. Files: Directly uploaded documents (PDF, DOCX, TXT, etc.)

3. Google Drive: Synced files from connected Google Drive accounts

All content is processed and indexed in Azure AI Search for consistent retrieval across agents.

Web Pages Knowledge Base

Dashboard Overview

The Web Page Crawler dashboard displays:

Limit: Maximum 100 URLs per agent

Crawling Strategies

• Depth-First: Explores each branch completely before moving to next

• Breadth-First: Explores all pages at current level before going deeper

Error Reporting

When crawling fails, the dashboard displays error details:

Common errors:

• Unreachable pages (404, 500 errors)

• Connection timeouts

• Parsing failures

• Strategy-related errors

• Access denied (authentication required)

Action Controls

Each URL entry provides three actions:

Resync

• Re-crawls the URL on demand

• Updates processed content and size

• Refreshes timestamps automatically

Download

• Downloads processed crawl output

• Enables offline review of extracted content

Delete

• Permanently removes URL and all associated data

• Removes from all integrated systems including Azure AI Search

Warning: Deletion is permanent and cannot be undone.

Files Knowledge Base

Dashboard Overview

The Files dashboard displays:

Limit: Maximum 100 files per agent

Supported File Types

• Documents: PDF, DOC, DOCX

• Spreadsheets: XLS, XLSX, CSV

• Text: TXT, MD

• Presentations: PPT, PPTX

Maximum file size: 10MB per file (configurable by admin)

Action Controls

Download

• Downloads the uploaded file

• Enables offline access or review

• Original file format preserved

Delete

• Permanently removes file from system

• Removes all associated references

• Removes from Azure AI Search index

Warning: Deletion is permanent and cannot be undone.

Google Drive Knowledge Base

Dashboard Overview

The Google Drive dashboard displays files organized by connected user accounts:

Per-Account View:

• Connected user email

• Total files synced

• Individual file listings

Per-File Information:

Limit: Maximum 100 files per connected account

Multi-Account Support

• Each user's files shown separately

• Clear visibility per account

• Independent sync status tracking

• Per-user file limits

Synchronization

Automatic Sync:

• Files sync automatically when modified in Drive

• Status updates reflect current sync state

• Timestamps show last successful sync

Sync Statuses:

• Synced: File successfully indexed in Azure AI Search

• Pending: Waiting for sync

• In Progress: Currently syncing

• Failed: Sync error occurred

Adding Content

Web Pages

1. Navigate to agent edit screen → Knowledge tab

2. Click Add Web Page

3. Enter root URL and select crawling strategy

4. Click Start Crawl

5. Monitor status in dashboard

Files

1. Navigate to agent edit screen → Knowledge tab

2. Click Upload File

3. Select file(s) from computer

4. Click Upload

5. Monitor sync status in dashboard

Google Drive

1. Navigate to agent edit screen → Knowledge tab

2. Click Connect Google Drive

3. Authenticate with Google account

4. Files sync automatically

5. Monitor in dashboard

Updating Content

• Web Pages: Click Resync to re-crawl updated website content

• Files: Delete old file and upload new version

• Google Drive: Files update automatically when modified in Drive

Removing Content

Click Delete action to permanently remove content from any knowledge source.

Troubleshooting

Crawl Failures

Common causes:

• Website accessibility issues

• URL incorrect

• Website requires authentication

• Robots.txt restrictions

Solutions:

• Verify URL is publicly accessible

• Check website allows crawling

• Try different crawling strategy

File Sync Issues

Common causes:

• Unsupported file format

• File size exceeds limit

• Network connectivity

Solutions:

• Convert to supported format

• Compress large files

• Retry upload

Google Drive Sync Problems

Common causes:

• Authentication expired

• File permissions in Drive

• Account connection inactive

Solutions:

• Re-authenticate Google Drive

• Verify file sharing settings

• Reconnect account if needed

Content Limits

Note: Limits are configurable by administrators and may vary by subscription tier.

Support

For assistance with knowledge base management:

• Documentation: CampusMind Knowledge Base

• Live Chat: Available in dashboard (bottom right)

• Email: support@campusmind.ai