Hey Product Hunt 👋 I'm Lisarde, the maker of Maskify. Maskify started from a simple frustration: every time I needed to share a PDF containing sensitive information — contracts, invoices, bank statements, screenshots — the available tools either uploaded the entire file to a third-party server or simply drew a black rectangle over text that was still copy-pasteable underneath. Neither felt truly private. So I built Maskify with a different approach: * Your original PDF/DOCX never leaves your browser. Files stay locally stored in IndexedDB while you work; only extracted text is sent for detection. * PDF exports are fully rasterised — the original text bytes are physically removed, not just visually hidden. No selectable text layer remains underneath. * DOCX files are modified in place while preserving styles, tables, and images. Optional metadata cleanup is included for extra privacy. * Hybrid detection system: * AI-based semantic detection using the open-source [OpenAI Privacy Filter](https://github.com/openai/privac...) model * Validated regex detection for patterns AI models often miss: * IBANs (mod-97 validated) * Credit cards (Luhn checked) * Spanish DNI/NIE * Common API key prefixes * Custom reusable patterns for recurring sensitive values * Local-only reversal map generation — only you can unmask later. * Supports PDF, DOCX, TXT, and Markdown. A few honest notes: * The core flow is completely free and doesn't require signup. * The AI detection currently runs on a small VPS (4 cores, ~12 GB RAM dedicated to the model), so processing is CPU-bound: * small documents take a few seconds * larger files can take up to ~30 seconds * Requests are processed one at a time. If another detection is already running, you'll enter a visible queue so you always know your position. * There's also a regex-only mode if you want near-instant results without AI processing. * No third-party AI APIs or external processing services are involved in the detection pipeline. I'd genuinely love feedback — especially on: * missed detections * false positives * edge-case documents * whether the privacy guarantees feel understandable from the UI Try it with your own messy real-world documents: 👉 https://maskify.es And tell me where it breaks 🙏

Maskify

Mask personal data in TXT, PDFs and DOCX, in your browser

Mask personal data in TXT, PDFs and DOCX, in your browser