Knowledge Base

The Knowledge Base is the foundation of Fid, where you organize and manage all your documentation, code repositories, and other knowledge sources. This comprehensive guide covers everything you need to know about building and maintaining an effective knowledge base.

Source Types

Fid supports a wide variety of source types to accommodate different documentation needs:

Document Sources

  • PDF Documents: Upload PDFs up to 50MB with automatic text extraction and OCR

  • Spreadsheets: Import CSV and Excel files with structured data parsing

  • Text Sources: Add custom text content with full markdown support

Code Repositories

  • GitHub: Index public and private repositories with branch/tag selection

  • File Filtering: Include or exclude specific file paths and extensions

  • Automatic Updates: Refresh repositories to capture latest changes

Web Sources

  • Websites: Crawl entire websites with configurable depth

  • Web Pages: Extract content from single pages

  • Doxygen: Import Doxygen-generated HTML documentation

  • Technical Docs: Support for Sphinx and other documentation frameworks

Enterprise Integrations

  • Confluence: Sync spaces from Atlassian Confluence

  • SharePoint: Connect to Microsoft SharePoint document libraries

  • External Syncs: Automated synchronization with regular updates

Adding Sources

Using the Web Interface

  1. Navigate to your project’s Knowledge Base tab

  2. Click the “Add Source” button

  3. Select your source type from the dropdown menu

  4. Fill in the required information:

    • For Documents: Upload file and provide a descriptive name

    • For GitHub: Enter repository URL, select branch/tag, configure filters

    • For Websites: Enter URL, set crawl depth, specify domain restrictions

    • For Enterprise: Authenticate and select spaces/libraries to sync

  5. Click “Add Source” to begin processing

Processing and Indexing

After adding a source, Fid will:

  1. Extract and process the content

  2. Generate embeddings for semantic search

  3. Create searchable indexes

  4. Extract metadata and structure

Processing time varies by source size and type:

  • Documents: Usually 1-2 minutes

  • GitHub repos: 2-10 minutes depending on size

  • Websites: 5-20 minutes based on crawl depth

  • Enterprise syncs: 10-30 minutes for initial sync

Source Organization

Groups and Hierarchies

Organize your sources effectively using groups:

  1. Create Groups: Click “New Group” to create a folder

  2. Nested Structure: Create subgroups for deeper organization

  3. Drag and Drop: Move sources between groups easily

  4. Bulk Operations: Select multiple sources for batch actions

Best Practices for Organization

  • By Topic: Group related documentation together

  • By Version: Separate different software versions

  • By Team: Organize by responsible teams or departments

  • By Type: Group similar source types together

Naming Conventions

Use clear, consistent naming:

  • Include version numbers: API Docs v2.1.0

  • Add dates for time-sensitive content: Q4 2024 Report

  • Use prefixes for categorization: [Backend] Authentication Module

  • Be descriptive but concise: Customer API Reference not docs

Advanced Features

Source Citations

Every chat response includes citations:

  • Numbered references link to source material

  • Click citations to view original content

  • Multiple citations show comprehensive coverage

  • Citations include page numbers and excerpts

Filtering and Targeting

Focus your searches:

  • Use @source-name to search specific sources

  • Use @group-name to search within groups

  • Combine multiple filters with commas

  • Exclude sources with -@source-name

Annotations

Add context to your documents:

  1. Open a PDF or document source

  2. Select text to annotate

  3. Add your note or clarification

  4. Annotations are searchable and appear in chat responses

Collaboration Features

Team Collaboration

Work together on knowledge bases:

  1. Shared Projects: All team members see the same sources

  2. Real-time Updates: Changes sync automatically

  3. Activity Tracking: See who added or modified sources

  4. Comments: Leave notes for team members

Access Control

Manage who can do what:

  • Viewers: Can search and chat, cannot modify

  • Editors: Can add and manage sources

  • Admins: Full control including settings and permissions

Version Control

Track changes over time:

  • Source version history

  • Rollback to previous versions

  • Compare changes between updates

  • Audit trail of modifications

Performance Optimization

Large Knowledge Bases

Tips for managing extensive content:

  1. Use Groups: Organize sources logically to improve navigation

  2. Regular Cleanup: Remove duplicate or obsolete sources

Search Performance

Optimize search speed:

  • Use targeted searches with @ mentions

  • Create focused project for specialized topics

Troubleshooting

Common Issues and Solutions

Source Won’t Process:

  • Check file size limits (50MB for PDFs)

  • Verify URLs are accessible

  • Ensure proper permissions for private repos

  • Check network connectivity

  • Refresh page after adding source

Search Not Finding Content:

  • Try tagging sources with @ mentions

  • Wait for indexing to complete

  • Try alternative search terms

Sync Failures:

  • Re-authenticate with external service

  • Check API rate limits

  • Verify permissions haven’t changed

  • Review sync logs for specific errors

Performance Issues:

  • Reduce number of active sources

  • Clear browser cache

  • Check internet connection speed

  • Contact support for large-scale issues

Next Steps

With your knowledge base established:

For help with specific source types or advanced features, please reach out to contact@fidlabs.ai.