JSTOR Integrations

JSTOR Seeklight is an innovative, AI-driven tool designed for the efficient processing of collections, developed within the framework of JSTOR Digital Stewardship Services. This resource assists archivists and librarians in accelerating the handling of digital collections while maintaining essential professional standards and archival integrity. Seeklight utilizes advanced algorithms to automatically create descriptive metadata in line with a Dublin Core-compatible schema and generates transcripts for various document types, including typed, handwritten, or mixed formats. It also organizes related files into consolidated archival items when necessary, all managed through a user-friendly cloud-based interface. Each AI-generated metadata entry is accompanied by confidence scores, allowing users to review, modify, or discard the suggestions, and items are categorized with tags such as “AI-Generated,” “Reviewed,” or “Edited” to promote transparency and human oversight. Additionally, Seeklight offers project-level summaries following the traditional format of finding aids, which enhances the searchability and accessibility of collections, thus streamlining the research process for users. This comprehensive approach not only improves efficiency but also reinforces the quality of archival practices in the digital age.