You are the last pair of eyes before a file leaves the building, and the file is usually a PDF full of names, a screenshot with a face in it, or a document whose hidden metadata says more than its pages do. Lolly cleans it on the device it is already on, with no upload, no account and no server in the middle.
Nothing leaves to be cleaned
- Black out, for good. Redact rebuilds the file around the boxes you draw, so the words under them are gone rather than covered, and it verifies the output before it downloads. PDFs and images alike.
- Anonymise the text. The de-identify mode in Text Helper swaps the personal identifiers in a block of text for placeholders, so an interview, a log or an export can be shared without the people in it.
- Strip what the pages do not show. Strip Hidden Data reveals and removes the metadata riding inside images and PDFs: locations, devices, authors, edit history.
- Shrink without sending. Compress recompresses a PDF's images on the device. The "free online compressor" is the exfiltration your data-loss policy exists to prevent.
- Check what you were sent. Verify reads a file's Content Credential where you are: who made it, with what, and whether AI was involved. The walk-through is Verify it yourself.
Every one of these is a utility in the app, listed together under Utilities, and every one of them works with the network cable pulled.
The positions, stated plainly
- Licence. MPL-2.0, with no contributor agreement. What that means for an organisation is in Adoption & Governance.
- AI marking. What Lolly declares about AI-assisted files, how, and what it does not claim, in AI marking and the EU AI Act.
- Data. What a server sees, which is nothing, in Privacy and Trust.