How it works
Four steps from student file to AI-ready
Everything happens in your browser. Your files, your token map and your restore key stay on your computer.
Set up a protected vault
Create a vault for a project, a course or a cohort and choose a passphrase. Import a roster from a CSV export so Blindfold Edu knows the names, nicknames and IDs to look for. The vault is encrypted on your device.
Load your file
Bring in a Word document, a text-based PDF, pasted text or survey comments. Names, student IDs, emails, phone numbers, street addresses, birthdates and school names are found and replaced with consistent tokens such as [STUDENT_014].
Review before anything leaves
A review screen shows every change, labeled by type. Sensitive terms such as accommodations or conduct notes are flagged for your decision rather than changed automatically. Click anything it missed to remove it, and Blindfold Edu remembers it for next time.
Use AI, then restore
Copy the clean version into Claude, ChatGPT, Copilot or any tool your campus allows. When the answer comes back, paste it into Blindfold Edu and your restore key puts the real names back, using the name each student goes by.
What it catches
Identifiers out, analysis in
REPLACED WITH TOKENS
Direct identifiers
Student names and nicknames, parent and staff names, student IDs, emails, phone numbers, street addresses and birthdates.
FLAGGED FOR YOU
Sensitive context
Accommodation, IEP and 504 references, conduct and Title IX language, and other terms that deserve a human decision.
KEPT EXACTLY
Your analysis fields
Credits, GPA, terms, course and program codes, scores and the text of the work itself.
COMING DECEMBER 2026
Spreadsheet mode for IR
Tokenize CSV and Excel extracts column by column, built for student-level files with thousands of rows.