Naosuu

On-device prompt redaction for AI platforms

Naosuu protects your sensitive data when you use AI platforms like ChatGPT, Claude, and Gemini. Naosuu scans your prompt for personally identifiable information such as names, email addresses, phone numbers, Social Security numbers, dates of birth, financial identifiers, legal case references, etc. Sensitive values are replaced with anonymous placeholders like [PERSON_1] or [EMAIL_1]. The AI never sees your real data. When the AI responds using those placeholders, Naosuu silently restores the original values in your browser so you read a natural response while your data stayed private the whole time. EVERYTHING RUNS ON YOUR DEVICE Naosuu uses on-device machine learning (NER) to detect PII. No prompt text, no document contents, and no detection results are ever sent to Naosuu or any third party. This is an architectural guarantee, not a policy choice. KEY FEATURES PII detection across 15+ categories including names, emails, phone numbers, SSNs, dates, financial identifiers, legal references, immigration documents, and more Review and confirm redactions via the preflight panel before sending Upload and redact sensitive PDFs and Word documents before attaching them to prompts Works on ChatGPT, Claude, and Gemini Response rehydration: original values are restored locally after the AI responds WHO IT IS FOR Lawyers, compliance officers, HR professionals, healthcare workers, and anyone who handles sensitive information and uses AI tools in their work. SUBSCRIPTION Naosuu requires a paid subscription. A free trial is available.