I built BlankOut because I kept running into the same hesitation that whenever I wanted to use AI to understand complex documents like medical reports, insurance forms, tax letters, invoices, or utility bills, I often gave up the attempt.
The problem was not that AI was not useful. It was that the documents often contained private names, addresses, IDs, account numbers, phone and other private data. Uploading those directly to a cloud AI tool felt uncomfortable and beyond our control.
So I started building an fully offline redaction tool to help redact private information locally before sharing documents with AI. It's on android playstore.
The goal is simple:
make AI more useful for real-life documents without uploading everything as-is. It is still early, and I would love feedback from other indie hackers! Appreciate any comments and suggestions.
Do you also avoid using AI when a document contains sensitive personal data (is this pain real for you as well?) what solutions did you try?
It is a massive barrier to productivity when you want to use AI for a quick summary but end up spending more time manually scrubbing the document for sensitive info than the task is actually worth.
The real security win here is the local processing because it ensures the data never leaves the device's RAM during the redaction phase which is the only way to truly satisfy strict privacy requirements.
Are you planning to add a batch processing feature so people can clean multiple files at once before feeding them into a bulk analysis workflow?
Yes, to extend the support parallel run of multiple files and pages is definitely in the roadmap. Thanks for resonating with the pain I am trying to address.