Getting Started with MaskIt
Redact PII from PDFs, SQL dumps, CSVs, and more — with multilingual NLP, multi-file batch upload, human review, and zero server-side processing.
Upload Files or Paste Text
Drop files or paste text directly into the dashboard. MaskIt accepts 6 file formats and supports multi-file batch uploads with live per-file progress tracking and automatic quota checking.
- PDF Documents— text extracted in-browser
- Structured Data: SQL, CSV, JSON
- Text & Logs: .txt and .log files
- Multi-file batch: drop several files at once
Configure PII Types & Thresholds
Fine-tune what MaskIt looks for. Toggle individual PII categories on or off, and adjust NLP confidence thresholds per entity type to balance precision and recall.
- Category toggles: enable or disable each of the 14 PII types independently.
- Confidence thresholds: adjust NLP minimum confidence for PERSON, ORG, and LOCATION.
- Pattern-only categories: regex types (email, IP, SSN, IBAN, etc.) match with 100% precision.
- Policy presets: save named configurations for different use cases (paid plans).
PII Category Toggles
NLP Confidence Thresholds
Smart Detection Engine
MaskIt combines deterministic regex patterns with a multilingual BERT-based NLP model to detect 14 PII categories across English and French content — all running in your browser.
Regex Engine (14 categories)
Email, IP, Credit Card, SSN (US & FR), Phone (US & FR), UUID, IBAN, URL, SIRET, SIREN, and DOB — deterministic and instant.
AI + Regex Hybrid (EN & FR)
Adds PERSON, ORG, and LOCATION via a multilingual BERT model. Runs in a dedicated Web Worker with configurable confidence thresholds and automatic retry/fallback.
Human Review & Audit
Every detection can be reviewed before finalization. Approve, reject, or re-categorize entities — then manually flag any PII the engine missed and fine-tune boundaries before exporting a complete audit trail.
- Triage controls: approve, reject, or re-categorize each entity with a single click.
- Manual flagging: select missed PII directly in the document, assign a category, and nudge boundaries for precision.
- File filtering: when reviewing a batch, filter by source file to focus on one document at a time.
- Batch actions & shortcuts: approve or reject all entities in a category. Keyboard shortcuts (A/R/D, arrows) for fast triage.
- Audit export: download the full decision log as JSON or CSV — metadata only, never raw PII.
Download Redacted Results
Export your redacted files in the same format as the original — or as a unified plain text file. Batch downloads are supported for multi-file jobs.
- Format-preserving: CSV, SQL, JSON, and TXT files keep their original structure.
- PDF exports as .txt: since redaction removes layout, PDF results are exported as plain text.
- Batch download: download all redacted files individually or as a single unified file.
Redacted output files
Analytics & History
Track every redaction run with real-time analytics. Global stats update automatically after each scan — no manual refresh needed.
- Global statistics: totals across all runs, not just the current page of results.
- Real-time updates: new runs appear immediately after redaction completes.
- Per-category breakdown: see which PII types appear most frequently.
Scan volume (7 days)
Total Scans
1,248
PII Found
14.2k
CLI Automation
Coming SoonTeam / EntIntegrate redaction into your CI/CD pipelines and local workflows. The MaskIt CLI allows developers to sanitize logs, CSVs, and PDFs before they ever leave their machine.
Found 128 PII entities across 6 categories.
✓ Redacted 14 files to ./redacted/logs
Public API
Coming SoonPro / Team / EntProgrammatically sanitize data streams from your backend services. A high-performance REST API ensures your internal tools and third-party integrations never handle raw PII.
Enterprise-grade reliability. Client libraries coming for Node.js, Python, and Go.