AI Features¶
Learn how to use Archivus's AI-powered features for document intelligence.
Quick Links¶
- Chat with Documents - Ask questions and get answers with citations
- Document Analysis - Summarization, entity extraction, classification
- Auto-Tagging - Automatic organization with AI
Getting Started¶
- Chat with Your First Document - Quick start guide
- Try Semantic Search - Find documents by meaning
- Learn about Auto-Tagging - Automatic organization
- Document Analysis - Advanced AI features
AI Capabilities¶
Chat (Archie)¶
Ask questions about your documents and get instant answers:
- Natural language questions
- Source citations and confidence scores
- Multi-document conversations
- Context-aware follow-ups
Semantic Search¶
Find documents by meaning, not just keywords:
- Understands intent and context
- Finds related concepts
- 10-100x better than keyword search
- Natural language queries
Auto-Tagging¶
Automatically tag and categorize documents:
- AI-suggested tags based on content
- Document type classification
- Entity extraction (people, dates, amounts)
- Smart categorization
Learn more about Auto-Tagging →
Document Analysis¶
Extract insights from your documents:
- Summarization - Quick and full summaries
- Entity Extraction - People, organizations, dates, amounts
- Classification - Automatic categorization
- Duplicate Detection - Find similar documents
How AI Processing Works¶
When you upload a document with AI enabled:
- Text Extraction - Extract text from document
- Content Analysis - Understand document meaning
- Entity Recognition - Find people, dates, organizations, amounts
- Summarization - Generate quick summary
- Tag Suggestions - Suggest relevant tags
- Semantic Indexing - Enable semantic search
Processing Time
- Small documents (< 10 pages): 30-60 seconds
- Medium documents (10-50 pages): 1-3 minutes
- Large documents (50+ pages): 3-10 minutes
AI Credit Usage¶
Different AI features use different amounts of credits:
| Feature | Credits | Frequency |
|---|---|---|
| Document Upload (with AI) | 1 | Per document |
| Chat Message | 1 | Per message |
| Quick Summary | 1 | Per request |
| Full Summary | 2 | Per request |
| Entity Extraction | 2 | Per request |
| Duplicate Detection | 1 | Per request |
Free Plan
Free plan includes 10 AI credits per month. Upgrade for more:
- Starter: 500 credits/month
- Pro: 1,500 credits/month
- Team: 6,000 credits/month
- Enterprise: Unlimited
Best Practices¶
Using Chat¶
- Be specific - "What are the payment terms?" vs "Tell me about this"
- Use follow-ups - Archie remembers context
- Check sources - Click citations for verification
- Attach documents - Use
@mentions for precision
Semantic Search¶
- Natural language - Ask like you're talking
- Be specific - Include context and details
- Use filters - Combine with folder/tag filters
- Try variations - Rephrase if needed
Auto-Tagging¶
- Review suggestions - AI tags before applying
- Correct mistakes - AI learns from corrections
- Combine with manual - Use both AI and manual tags
- Be consistent - Use standard tag formats
Document Analysis¶
- Use quick summaries - For overviews
- Use full summaries - For detailed analysis
- Verify entities - Check confidence scores
- Review classifications - Verify AI categories
Common Use Cases¶
Contract Analysis¶
- Upload contract
- Ask Archie: "What are the key terms?"
- Review summary and entities
- Check for similar contracts
Invoice Processing¶
- Upload invoice (batch upload supported)
- AI extracts vendor, amount, date
- Auto-tags and categorizes
- Routes to correct folder
Document Research¶
- Use semantic search to find relevant documents
- Chat with results to extract insights
- Compare multiple documents
- Summarize findings
Compliance Review¶
- Upload policy documents
- AI classifies and tags
- Search for specific requirements
- Check for contradictions
Related Guides¶
- Document Management - Upload and organize documents
- Collaboration - Team collaboration features
Need Help?¶
- FAQ - Frequently asked questions
- Contact Support - Get help
- Ask Archie - "What can you help me with?"