Security & privacy

Browser-first PDF processing

Most Dr.PDF tools - merge, split, compress, sign, protect, OCR, and most edits - run entirely in your browser using JavaScript, WebAssembly, and the File API. When you drop a PDF, it stays in browser memory on your machine.

Office conversions

Word, Excel, and PowerPoint tools may upload your document to our server for conversion with LibreOffice when that engine is available. Files are processed in memory and discarded immediately - not stored in a database or shared drive. When LibreOffice is unavailable, the same tools may use a browser fallback without uploading your file.

What we don't collect

• Your PDF files, images, or Office documents
• File names or metadata from your uploads
• Account credentials - there are no accounts
• Processing history or tool usage tied to identity

HTTPS & transport

Dr.PDF is served over HTTPS, ensuring the application code itself is delivered securely. Because files are not uploaded, your document content is never in transit to our servers.

Encryption tools

Password protection uses RC4 128-bit encryption, the widely supported PDF standard handler. It is a solid deterrent against casual access, but it is not modern strong encryption - do not rely on it alone for highly sensitive material. Protect PDF encrypts locally before download. Unlock PDF verifies your password, then re-saves the document without encryption.

Redaction

Redact PDF rasterises each page and permanently discards the underlying text layer, so redacted content cannot be recovered by selecting or extracting text. Because this requires canvas rendering, the tool refuses to run rather than fall back to drawing black boxes over text that would remain readable.

Third-party services

Your PDF files are not stored on our servers. These tools do make network requests:

• Office conversions may upload documents temporarily for LibreOffice processing (see above).
• Translate PDF sends extracted text snippets (up to 500 characters per page) to MyMemory, a public translation API. The original file is never sent. This tool requires explicit confirmation before it runs.
• OCR PDF downloads language models from a public CDN on first use. No document content leaves your device - recognition itself runs locally.

All other PDF tools run fully in the browser after the page loads.

Compared to cloud PDF tools

Most cloud-based PDF services store and process files on their infrastructure. Dr.PDF minimizes server exposure: PDF tools stay on your device, and Office uploads are temporary and discarded after conversion.

Operations & compliance

Deployment, health checks, Docker setup, AdSense readiness, and go-live checklists are in deploy.md in the project repository. Extended ops topics are in docs/SAAS_OPERATIONS.md.

Recommendations for sensitive documents

• Use Dr.PDF on a trusted device and browser
• Close the tab after processing to clear browser memory
• For highly sensitive workflows, use a dedicated browser profile
• Verify downloads before sharing externally