ArchiveTriage

ArchiveTriage is a web app (with optional desktop uploader) that helps libraries, archives, and institutional repositories find and fix broken, inconsistent, or incomplete metadata at scale. It connects to common systems (DSpace, Fedora/Islandora, Omeka, S3-backed collections) and runs automated audits: missing required fields, invalid dates, duplicate identifiers, inconsistent controlled vocab terms, and suspicious OCR/text extraction gaps. An AI-assisted assistant proposes fixes (e.g., normalize names, suggest subjects from existing thesauri, flag probable duplicates), but every change is reviewable with an approval workflow and full change logs. The goal is not “magic AI cataloging”; it’s triage: quickly surfacing the worst problems, prioritizing what matters for discovery, and making safe, reversible bulk edits. Export reports for grant compliance and internal QA, and schedule recurring audits so metadata debt doesn’t silently grow.