Chat with PDF

Overview

Chat with PDF lets you upload any PDF document and have intelligent conversations about its content. Ask questions, extract information, and get insights without reading the entire document.

Supported Document Types

File Formats

• PDF (.pdf) - Standard and scanned
• Microsoft Word (.doc, .docx)
• Rich Text Format (.rtf)
• Plain Text (.txt)
• Markdown (.md)
• Microsoft Excel (.xlsx, .xls)
• Microsoft PowerPoint (.pptx, .ppt)
• OpenDocument (.odt, .ods, .odp)

File Size Limits

• Standard Plan: Up to 75MB per file
• Pro Plan: Up to 150MB per file
• Enterprise: Up to 500MB per file
• Recommended: Under 25MB for optimal performance

Document Quality

Best Results:

• ✅ Text-based PDFs (selectable text)
• ✅ Well-formatted documents
• ✅ Clear fonts and layout
• ✅ High-resolution scans (300+ DPI)
• ✅ Machine-printed text

Acceptable:

• ⚠️ Scanned documents (OCR applied)
• ⚠️ Mixed text and images
• ⚠️ Tables and charts
• ⚠️ Multi-column layouts

Challenging:

• ❌ Handwritten text
• ❌ Very low resolution scans
• ❌ Heavy image-based PDFs
• ❌ Corrupted or damaged files
• ❌ Password-protected PDFs

Uploading Documents

Upload Methods

Method 1: Drag and Drop

1. Navigate to Documents page
2. Drag PDF from your computer
3. Drop into upload area
4. Watch processing progress

Method 2: File Browser

1. Click "Upload Documents" button
2. Browse files on computer
3. Select one or more files
4. Click "Open"

Method 3: Cloud Import

• Import from Google Drive
• Import from Dropbox
• Import from OneDrive
• Sync folders automatically

Method 4: Email Upload

• Forward document to your unique upload email
• Format: uploads-[id]@solatis.team
• Automatically processed
• Notification when ready

Upload Process

What Happens:

1. Upload File → Validation → Security Scan
   ↓
2. Extract Text → Language Detection
   ↓
3. Process Content → AI Analysis → Generate Embeddings
   ↓
4. Index & Store → Ready for Chat

Processing Time:

• Small PDFs (< 10 pages): 30-60 seconds
• Medium PDFs (10-50 pages): 1-3 minutes
• Large PDFs (50+ pages): 3-10 minutes
• Scanned PDFs: Add 30-50% time for OCR

Progress Indicators:

• 🔄 Uploading: File transferring
• 📄 Extracting: Reading content
• 🤖 Processing: AI analysis
• ✅ Ready: Available for chat

Chatting with Documents

Starting a Conversation

Quick Start:

1. Upload document (or select existing)
2. Click document name
3. Click "Chat" tab
4. Ask your first question

Example First Questions:

What is this document about?
Summarize the main points
What are the key recommendations?
Who is the intended audience?

Document Analysis

Automatic Analysis:

Upon upload, AI generates:

• Executive Summary: 2-3 paragraph overview
• Key Points: Bullet list of main ideas
• Topics: Main themes and subjects
• Entities: People, organizations, locations, dates
• Sentiment: Overall tone (positive, neutral, negative)
• Document Type: Report, contract, article, etc.

View Analysis:

1. Open document
2. Click "Intelligence" or "Analysis" tab
3. See AI-generated insights

Question Types

Factual Questions:

What is the contract start date?
Who signed the agreement?
What is the total budget?
How many participants were there?

Analytical Questions:

What are the main risks identified?
How does this compare to previous reports?
What patterns emerge from the data?
What are the root causes mentioned?

Extraction Requests:

List all action items
Extract all dates and deadlines
Find all financial figures
Pull out all recommendations

Synthesis Questions:

Summarize each section
What are the key takeaways?
Create an executive brief
Identify the main arguments

Advanced Querying

Page-Specific:

What does page 5 discuss?
Summarize pages 10-15
What's in the introduction section?

Comparative:

Compare Section 2 and Section 4
How do the conclusions differ from the introduction?
What changed between version 1 and version 2?

Topic-Focused:

What does the document say about cybersecurity?
Find all mentions of budget constraints
Where is the implementation timeline discussed?

Document Intelligence Features

Semantic Search

How It Works:

• Understands meaning, not just keywords
• Finds relevant content even without exact matches
• Context-aware matching
• Ranked by relevance

Examples:

Search: "financial performance"
Finds:
- Revenue and profit discussions
- Balance sheet information
- Cost analysis sections
- ROI calculations

Entity Extraction

Automatically Identifies:

People:

• Names and roles
• Contact information
• Mentions and references

Organizations:

• Company names
• Department references
• Partner organizations

Locations:

• Addresses
• Cities, states, countries
• Office locations

Dates & Times:

• Deadlines
• Meeting dates
• Project timelines
• Historical references

Financial:

• Currency amounts
• Budget figures
• Costs and revenues
• Financial metrics

Example Output:

People Mentioned:
- Sarah Johnson (CEO)
- John Smith (CFO)
- Alice Chen (Project Lead)

Organizations:
- Acme Corp (client)
- TechVendor Inc (partner)

Key Dates:
- Project Start: Nov 1, 2024
- Milestone 1: Dec 15, 2024
- Final Delivery: Feb 28, 2025

Table Extraction

Capability:

• Detect tables in documents
• Extract tabular data
• Convert to structured format
• Export as CSV or Excel

Example:

Extract budget table from page 12

Result:
| Category | Q3 2024 | Q4 2024 | Change |
|----------|---------|---------|--------|
| Personnel| $450K   | $475K   | +5.6%  |
| Marketing| $200K   | $180K   | -10%   |
| R&D      | $300K   | $325K   | +8.3%  |

Document Comparison

Compare Multiple Documents:

1. Select 2-3 related documents
2. Click "Compare Documents"
3. AI analyzes differences

Comparison Types:

• Version Comparison: Track changes between versions
• Content Comparison: Different documents on same topic
• Temporal Comparison: Documents across time periods
• Format Comparison: Same content, different formats

Example:

Compare Contract v1.0 and Contract v2.0

Changes Found:
- Payment terms updated (Net 30 → Net 45)
- Added Clause 7: Intellectual Property
- Modified Clause 3.2: Liability limits increased
- Removed Clause 9: Non-compete provision

Use Cases

Legal Documents

Contract Review:

You: What are the key terms of this contract?

AI: The contract includes:

1. **Parties**: TechCo (Provider) and Acme Corp (Client)
2. **Term**: 24 months starting Jan 1, 2025
3. **Value**: $500,000 annual fee
4. **Payment**: Quarterly invoices, Net 30
5. **Services**: Cloud hosting and support
6. **Termination**: 60-day written notice
7. **Liability**: Limited to annual fees
8. **Renewal**: Auto-renew unless cancelled

Key Clauses:
- Section 4.2: Data security requirements
- Section 6.1: Service level agreements (99.9% uptime)
- Section 8.3: Intellectual property ownership

[Citations with page numbers provided]

Risk Assessment:

You: What potential risks are in this contract?

AI: Identified risks:

HIGH RISK:
- Automatic renewal clause without cap (Sec 2.3)
- Unlimited liability for data breaches (Sec 7.4)
- One-sided termination fees (Sec 10.2)

MEDIUM RISK:
- Vague performance metrics (Sec 5.1)
- No disaster recovery timeline specified
- Missing force majeure provisions

LOW RISK:
- Standard confidentiality terms
- Reasonable payment schedule
- Clear scope definition

Recommendations:
1. Negotiate liability cap
2. Add specific SLA metrics
3. Clarify auto-renewal terms

Financial Reports

Key Metrics Extraction:

You: What are the key financial metrics?

AI: Q3 2024 Financial Highlights:

Revenue:
- Total Revenue: $2.5M (+15% YoY)
- Recurring Revenue: $1.8M (72% of total)
- New Business: $700K

Profitability:
- Gross Margin: 68% (target: 65%)
- Operating Income: $450K (+22% YoY)
- Net Profit Margin: 18%

Cash Flow:
- Operating Cash Flow: $520K
- Free Cash Flow: $380K
- Cash on Hand: $1.2M

Key Ratios:
- Customer Acquisition Cost: $850
- Lifetime Value: $12,400
- LTV/CAC Ratio: 14.6x

Source: Q3 Financial Report, pages 3-7

Research Papers

Literature Review:

You: What methodology was used in this study?

AI: Research Methodology:

Study Design:
- Type: Randomized controlled trial
- Duration: 12 months
- Sample Size: 250 participants (125 per group)
- Setting: Multi-site (5 locations)

Data Collection:
- Surveys (baseline, 3, 6, 12 months)
- Biomarker measurements
- Activity tracking (wearable devices)
- Qualitative interviews (n=50)

Statistical Analysis:
- Primary: Intent-to-treat analysis
- Methods: Mixed-effects models, ANOVA
- Software: R version 4.2, SPSS 28
- Significance level: p < 0.05

Limitations:
- Self-reported data (potential bias)
- High dropout rate (18%)
- Limited generalizability (urban population only)

Technical Documentation

Implementation Guide:

You: How do I implement the authentication system?

AI: Authentication Implementation Steps:

Prerequisites:
- Node.js 18+
- PostgreSQL 14+
- Redis 7+

Step 1: Install Dependencies
```bash
npm install jsonwebtoken bcrypt express-session

Step 2: Configure Environment

JWT_SECRET=your_secret_key
SESSION_SECRET=your_session_secret
DATABASE_URL=postgresql://localhost:5432/app

Step 3: Set Up Middleware [Code example from pages 15-17]

Step 4: Create Auth Routes [Code example from pages 18-20]

Step 5: Implement Password Hashing [Code example from page 22]

Testing:

• Unit tests: pages 45-48
• Integration tests: pages 49-52

Security Notes:

• Use HTTPS in production
• Rotate secrets regularly
• Implement rate limiting
• Enable 2FA for sensitive operations

### Meeting Minutes

**Action Item Extraction:**

You: What are the action items from this meeting?

AI: Action Items from Product Planning Meeting:

High Priority:

1. Update pricing page with new tiers

   • Owner: Sarah
   • Due: Nov 15, 2024
   • Status: Not started

2. Schedule customer interviews (n=10)

   • Owner: Marketing team
   • Due: Nov 30, 2024
   • Status: In progress

3. Review competitor pricing

   • Owner: John
   • Due: Nov 10, 2024
   • Status: Not started

Medium Priority: 4. Draft feature comparison matrix

• Owner: Alice
• Due: Nov 20, 2024

5. Create mockups for new features
   • Owner: Design team
   • Due: Dec 1, 2024

Decisions Made:

• Approved $50K additional marketing budget
• Selected Vendor A for hosting migration
• Postponed mobile app launch to Q1 2025

Next Meeting: Nov 12, 2024 at 2:00 PM

## Export and Sharing

### Export Options

**Export Entire Document:**
1. Open document
2. Click "..." menu → "Export"
3. Choose format:
   - Original PDF
   - Markdown (text only)
   - Word (DOCX)
   - JSON (with metadata)
4. Download

**Export Chat History:**
1. In chat view
2. Click "Export Conversation"
3. Formats:
   - PDF (formatted conversation)
   - Markdown (plain text)
   - JSON (structured data)

**Export Insights:**
1. Go to Intelligence tab
2. Click "Export Analysis"
3. Includes:
   - Summary
   - Key points
   - Entities
   - Topics
   - Action items

### Sharing Documents

**Share with Team:**
1. Click "Share" button
2. Select team members
3. Set permissions:
   - View only
   - Can comment
   - Can edit
   - Can reshare
4. Optional: Add message
5. Click "Send"

**Generate Share Link:**
1. Click "Share" → "Get Link"
2. Options:
   - Anyone with link
   - Organization only
   - Specific people
3. Set expiration (optional)
4. Password protect (optional)
5. Copy link

**Email Sharing:**
1. Click "Share" → "Email"
2. Enter recipient emails
3. Add message (optional)
4. Choose to include:
   - Original document
   - AI summary
   - Chat history
5. Send

## Document Management

### Organization

**Folders:**
- Create folder structure
- Drag documents to folders
- Nested folders supported
- Shared team folders

**Tags:**
- Add multiple tags per document
- Color-coded organization
- Filter by tags
- Suggested tags (AI-powered)

**Collections:**
- Smart collections (auto-updating)
- Manual collections
- Shared collections
- Quick access

**Metadata:**
- Title
- Description
- Author
- Date
- Custom fields
- Version number

### Search and Filter

**Search Options:**
- Full-text search
- Semantic search
- Filter by type
- Filter by date
- Filter by owner
- Filter by tags

**Advanced Filters:**

type:pdf AND tag:contract AND date:2024 author:"john@company.com" NOT archived:true text:"quarterly report" AND size:>5MB

### Version Control

**Track Changes:**
- Automatic version history
- Compare versions
- Restore previous versions
- See who made changes

**Version Notes:**
- Add version descriptions
- Track major vs minor changes
- Link related versions
- Merge annotations

## Best Practices

### Upload Tips

**Do:**
- ✅ Use descriptive file names
- ✅ Add tags immediately after upload
- ✅ Organize into folders
- ✅ Add descriptions for context
- ✅ Verify OCR quality for scans

**Don't:**
- ❌ Upload password-protected PDFs (remove protection first)
- ❌ Use generic names ("document1.pdf")
- ❌ Skip metadata
- ❌ Upload extremely large files (split if possible)
- ❌ Upload copyrighted material without permission

### Effective Questions

**Do:**
- ✅ Be specific about what you need
- ✅ Reference page numbers if known
- ✅ Ask follow-up questions
- ✅ Request examples and details
- ✅ Verify AI responses with source

**Don't:**
- ❌ Ask overly broad questions
- ❌ Expect perfect accuracy without verification
- ❌ Rely solely on AI for critical decisions
- ❌ Skip reading important passages
- ❌ Forget to check citations

## Troubleshooting

**Upload Fails:**
- Check file size (under limit)
- Verify file format supported
- Ensure stable internet connection
- Try different browser
- Remove password protection

**Poor OCR Quality:**
- Re-scan at higher resolution (300+ DPI)
- Ensure clear, readable text
- Try different file
- Manual data entry if needed

**Inaccurate Responses:**
- Verify source document quality
- Check if document fully processed
- Ask more specific questions
- Cross-reference with original
- Report persistent issues

**Slow Processing:**
- Normal for large files
- Check system status
- Try during off-peak hours
- Split large documents
- Contact support if > 30 min

## Next Steps

- [Chat with AI](/features/chat-with-ai) - Advanced conversation techniques
- [Audio Transcriber](/features/audio-video-transcriber) - Transcribe meetings
- [Document Intelligence](/features/business-intelligence) - Analytics and insights
- [Integrations](/integrations/google-workspace) - Connect cloud storage

---

**Need help?** Check [Troubleshooting](/troubleshooting/common-issues) or [contact support](/troubleshooting/contact-support).

**Last Updated:** October 11, 2025