Drop in a batch of resumes and BLIND removes the things that trigger unconscious bias, the name, the contact details, the home address, the personal links, the graduation years that hint at age, the photo and the gendered pronouns, and hands back clean packets labelled Candidate #1, #2, and so on. A separate key maps the numbers back to the files, so your reviewers stay blind and you lose nothing. Included in the Sovereign Pass.
Blinded packets appear here.
The reviewers get the packets. The person running the process keeps the key, which maps Candidate #N back to the source file and the name that was detected, so a shortlist can be un-blinded when the review is done.
It removes the markers, consistently. The candidate's name (detected from the top of the resume), emails, phones, addresses, personal and social links, graduation years, and gendered pronouns and titles. Photos come out because the text is extracted and the images are left behind.
It standardises to text on purpose. A blinded packet is plain text, not a formatted resume. That is deliberate: formatting, fonts and photos are themselves identifying, and stripping them levels the field.
It is not perfect anonymity. A distinctive employer, a named project, an unusual phrasing or a rare qualification can still point to a person, and name detection is a heuristic that can miss an unusual name or a name only in the file. Read a couple of packets, and treat this as a strong first pass, not a legal guarantee of a bias-free process.
A scanned resume is a picture. If a PDF is a photo of a page rather than real text, no text can be extracted; ask for a text PDF or a .docx.
Nothing leaves this tab. Resumes are read and blinded in browser memory.