PowerPoint Redaction

PowerPoint presentations are ubiquitous in business communication—sales pitches include customer details, HR presentations contain employee information, financial decks display sensitive metrics, and training materials reference real cases. Unlike simpler document formats, presentations layer text across multiple elements: slide content, speaker notes, charts, tables, SmartArt, and embedded objects. Each element type requires specific handling to detect and redact PII effectively while preserving the visual design that makes presentations impactful.

Our PowerPoint processing comprehensively addresses every location where sensitive data might exist. We parse the complex PPTX structure to access all text-containing elements, apply PII detection uniformly, and maintain the presentation's visual integrity through redaction. The result is a protected presentation that looks and functions exactly like the original—minus the sensitive information.

Presentation Structure

PPTX files are ZIP archives containing XML documents:

presentation.pptx/
├── [Content_Types].xml
├── _rels/
├── ppt/
│   ├── presentation.xml
│   ├── slides/
│   │   ├── slide1.xml
│   │   ├── slide2.xml
│   │   └── ...
│   ├── slideMasters/
│   ├── slideLayouts/
│   ├── notesSlides/
│   ├── charts/
│   ├── embeddings/
│   ├── media/
│   └── theme/
└── docProps/
    ├── core.xml
    └── app.xml

Processing Order:

• Document properties (metadata)
• Slide masters and layouts (template content)
• Individual slides (main content)
• Notes slides (speaker notes)
• Charts and embedded content
• Media files (images via OCR)

Slide Content Redaction

Slide content exists in various containers:

Text Boxes and Shapes:

// Typical slide text structure
<p:sp>  // Shape
  <p:txBody>  // Text body
    <a:p>  // Paragraph
      <a:r>  // Run
        <a:rPr>...</a:rPr>  // Run properties (formatting)
        <a:t>Contact: [email protected]</a:t>  // Text
      </a:r>
    </a:p>
  </p:txBody>
</p:sp>

// After redaction
<a:t>Contact: [EMAIL]</a:t>

// Formatting (font, size, color) preserved

Placeholder Text:

• Title placeholders
• Content placeholders
• Subtitle placeholders
• Footer, date, and slide number fields

Smart Text Features:

// Hyperlinked text
<a:r>
  <a:rPr>
    <a:hlinkClick r:id="rId1"/>  // Link to [email protected]
  </a:rPr>
  <a:t>Contact John</a:t>
</a:r>

// Both display text and link URL are redacted
// Link can be removed or pointed to placeholder

Speaker Notes Processing

Speaker notes often contain additional sensitive details:

// Notes slide structure
ppt/notesSlides/notesSlide1.xml

// Notes may contain:
- Extended talking points with real names
- Customer-specific details for customization
- Confidential context not shown to audience
- References to specific contracts or deals

// Processing
1. Extract text from notes slides
2. Apply same PII detection as slide content
3. Redact while preserving note formatting
4. Optionally: Remove notes entirely for external versions

Notes Configuration:

{
  "processSpeakerNotes": true,
  "notesBehavior": "redact",  // or "remove" for external versions
  "preserveNotesFormatting": true
}

Charts and Data Visualization

Charts contain data that may include PII:

Embedded Charts:

// Chart data stored in embedded Excel workbook
ppt/embeddings/Microsoft_Excel_Worksheet1.xlsx

// Chart elements to process:
- Series names (might be customer names)
- Category labels (might be identifiers)
- Data labels on chart
- Chart title and axis titles
- Data table if displayed

// Processing approach:
1. Extract embedded workbook
2. Process as Excel file (all sheets)
3. Update chart cache in chart XML
4. Replace embedded workbook

Chart Configuration:

{
  "processCharts": true,
  "chartElements": {
    "seriesNames": true,
    "categoryLabels": true,
    "dataLabels": true,
    "titles": true,
    "embeddedData": true
  }
}

Tables in Presentations

Tables frequently contain structured PII:

// PowerPoint table structure
<a:tbl>
  <a:tr>  // Row
    <a:tc>  // Cell
      <a:txBody>
        <a:p>
          <a:r><a:t>Customer Name</a:t></a:r>
        </a:p>
      </a:txBody>
    </a:tc>
    <a:tc>
      <a:txBody>
        <a:p>
          <a:r><a:t>[email protected]</a:t></a:r>
        </a:p>
      </a:txBody>
    </a:tc>
  </a:tr>
</a:tbl>

// Cell-by-cell processing
// Header row context helps classify data
// Formatting (borders, colors, merges) preserved

SmartArt Graphics

SmartArt contains text in diagram elements:

// SmartArt structure
ppt/diagrams/
├── data1.xml      // Diagram data
├── drawing1.xml   // Visual representation
├── colors1.xml    // Color scheme
├── layout1.xml    // Layout definition
└── quickStyle1.xml

// Text in SmartArt
<dgm:pt modelId="1">
  <dgm:t>
    <a:p>
      <a:r><a:t>John Smith - CEO</a:t></a:r>
    </a:p>
  </dgm:t>
</dgm:pt>

// Organization charts often contain names/titles
// Process diagrams are usually safe
// Hierarchy diagrams may have names

Embedded Objects

Presentations often embed other documents:

OLE Objects:

// Embedded Excel worksheet
ppt/embeddings/Microsoft_Excel_Worksheet1.xlsx

// Embedded Word document
ppt/embeddings/Microsoft_Word_Document1.docx

// Embedded PDF (as image with OLE wrapper)
ppt/embeddings/oleObject1.bin

// Processing:
1. Identify embedded object type
2. Extract to temporary location
3. Process with appropriate handler
4. Replace embedded object
5. Update OLE object image preview

Linked Objects:

{
  "linkedObjects": {
    "processLinks": true,
    "behavior": "process_source",  // or "visible_only"
    "breakLinksAfter": false
  }
}

Image Processing

Images may contain text requiring redaction:

// Image locations
ppt/media/
├── image1.png   // Screenshots
├── image2.jpeg  // Photos
└── image3.emf   // Vector graphics

// OCR processing for images
1. Extract images from media folder
2. Apply OCR to detect text
3. Identify PII in extracted text
4. Apply visual redaction (black boxes)
5. Replace original image

// Configuration
{
  "processImages": true,
  "imageOCR": {
    "enabled": true,
    "languages": ["en"],
    "minConfidence": 0.7
  },
  "imageRedactionStyle": {
    "method": "blackbox",
    "color": "#000000"
  }
}

Master Slides and Layouts

Template elements may contain PII:

// Slide master content
ppt/slideMasters/slideMaster1.xml

// Elements that might have PII:
- Company name/logo (usually ok)
- Contact information in footer
- Presenter name in layout
- Template sample text

// Processing masters affects all slides using them
// Can selectively process masters or skip

Metadata and Properties

Document properties contain sensitive information:

// Core properties
docProps/core.xml
- dc:creator (Author)
- cp:lastModifiedBy
- dc:title
- dc:subject
- cp:keywords

// Extended properties
docProps/app.xml
- Company
- Manager
- PresentationFormat

// Custom properties
docProps/custom.xml
- User-defined properties

// Redaction options
{
  "metadata": {
    "removeAuthor": true,
    "removeLastModifiedBy": true,
    "removeCompany": true,
    "removeCustomProperties": true,
    "preserveTitle": true
  }
}

Comments and Annotations

Comments often contain sensitive discussions:

// Comment structure
ppt/comments/comment1.xml

<p:cm authorId="0" dt="2024-01-15T10:30:00Z">
  <p:text>
    Check with John Smith before presenting
    to Acme Corp - they have special pricing
  </p:text>
</p:cm>

// Options:
{
  "comments": {
    "process": true,     // Redact PII in comments
    "remove": false,     // Or remove all comments
    "removeAuthors": true  // Remove author attribution
  }
}

Animations and Transitions

Visual effects are preserved through redaction:

// Animation XML is separate from content
// Redacting text doesn't affect animation settings

// Preserved elements:
- Slide transitions
- Object entrance/exit animations
- Motion paths
- Animation timing and triggers
- Sound effects
- Morph transitions (PPTX 2019+)

// Text redaction maintains text box dimensions
// Animations targeting text boxes continue working

Password Protection

Handle protected presentations:

// Protection types:
1. Open password - Required to open file
2. Modify password - Required to edit
3. Mark as Final - Advisory, not enforced

// API configuration
{
  "file_url": "https://example.com/protected.pptx",
  "passwords": {
    "open": "secretPassword123",
    "modify": "editPassword456"
  },
  "outputProtection": {
    "retainProtection": false,
    "newOpenPassword": null,
    "newModifyPassword": null
  }
}

Legacy PPT Format

Support for Office 97-2003 format:

// PPT is binary format (CFBF container)
// Different internal structure than PPTX

// Processing approach:
1. Parse binary PPT structure
2. Extract text from records
3. Apply PII detection
4. Modify binary records
5. Preserve binary format (no conversion to PPTX)

// Supported PPT features:
- All text content
- Speaker notes
- Embedded objects
- Document properties
- Comments

// Some limitations vs PPTX:
- More complex binary parsing
- Less granular control

API Usage

POST /v1/redact/file
Content-Type: multipart/form-data

{
  "file": [binary PPTX data],
  "filename": "sales_presentation.pptx",
  "options": {
    "redaction_types": ["pii", "financial"],
    "processElements": {
      "slides": true,
      "speakerNotes": true,
      "charts": true,
      "tables": true,
      "smartArt": true,
      "embeddedObjects": true,
      "images": true,
      "comments": true,
      "metadata": true
    },
    "preserveFormatting": true,
    "outputFormat": "pptx"
  }
}

Response:
{
  "job_id": "job_abc123",
  "status": "completed",
  "download_url": "https://api.redactionapi.net/v1/download/xyz",
  "statistics": {
    "slides_processed": 24,
    "elements_scanned": 156,
    "detections": {
      "email": 12,
      "name": 8,
      "phone": 3
    },
    "processing_time_ms": 2850
  }
}