Data & confidentiality

On servers located in India, the EU, and the USA. Manuscripts uploaded to QC by Galleys and journal databases we host are kept on separate systems, and each customer’s data is isolated — no publisher can see another’s. All connections use HTTPS.

No. Manuscripts and journal data are never used to train, fine-tune, or evaluate any model, and are never sold, licensed, or shared. Some screening engines send manuscript text to a commercial language-model API; under those providers’ terms, inputs are excluded from model training. We’ll name our current providers on request.

Manuscript files are deleted seven days after screening. Reports are kept for as long as your account is active, so you can compare a revision against the original run.

Duplicate detection works from a fingerprint, not the file — a non-reversible signature of the text plus basic metadata like title and authors. The original manuscript cannot be reconstructed from it. This is what lets us catch a resubmission months later without storing the paper itself. Your fingerprints are matched only against your own archive; they are never pooled with other publishers’ submissions.

You can delete any manuscript, report, or fingerprint yourself at any time, or email us and we’ll remove it within one business day.

Your data goes with you. We export your full report archive, or your complete journal database and files, on request and at no charge — during the relationship or after it. We keep nothing once handover is confirmed.

Yes, before any manuscript is uploaded. We can also execute a data processing agreement for universities and institutional publishers. Send us your standard form or ask for ours.

Email contact@galleys.pub or message us on WhatsApp. We’ll answer specific questions about infrastructure, subprocessors, backups, or access controls directly — and under NDA if you’d prefer.