De-ID.org
Browser-based HIPAA PHI De-Identification
About De-ID.org
De-ID.org is a browser-based HIPAA PHI de-identification engine built for healthcare AI teams, researchers, and privacy-first developers. Most AI tools require you to upload sensitive medical documents to their servers. That creates compliance risk, audit complexity, and trust issues. De-ID solves this differently. It runs ONNX-based NER models directly inside the browser using ONNX Runtime Web. No PHI leaves the device. No server storage. No retention. Upload a clinical PDF → extract structured text → detect all 18 HIPAA identifiers → download anonymized output as Markdown, JSON, or TXT. Key capabilities: • Client-side ML inference (true browser execution) • PDF structure preservation (headers, tables, formatting) • Entity confidence scores • Rule-based cascade for SSN, MRN, regex patterns • Zero infrastructure cost for privacy-first mode If you're building healthcare AI and need PHI stripped before sending data to LLMs, this is the missing layer
Category: ai_tool
Location: Mumbai, India