Archival Transcription & Finding Aids demo
Transcribes letters, ledgers and oral histories into DACS finding aids, for archivists 1:29
What the demo shows
Transcribes letters, ledgers and oral histories into DACS finding aids, for archivists
An Intake agent checks each capture batch by box and folder. A Transcription agent reads handwritten memos, typed carbons, ledgers, catalog cards and oral-history audio, giving every line a confidence; a Reader agent independently re-reads the uncertain lines, and where the two disagree the line is marked [illegible?] for a person. Fix a reading once and the agents propose it wherever the same hand appears. A Describer agent drafts DACS folder titles, dates and scope notes, citing the page behind every fact, each line a draft until an archivist approves it. A Sensitivity agent screens for FERPA student records, medical details and identifiers and applies deeds of gift and transfer agreements; it can only hold a folder. The processing archivist confirms a restriction, the head of special collections approves it, an Accessibility agent checks WCAG 2.1 AA, and a Publisher agent releases the box to the archival management system, the digital asset system and discovery.
Transcript
- Archival Transcription and Finding Aids turns scanned archives into cited finding aids, for archivists.
- Halden State's backlog is an estimated sixty-nine staff-years of work.
- Sixty-four boxes sit on this shelf; each fills as agents work.
- Overnight, agents read Box 18's 141 pages, scoring every line.
- Disagreements go to a person, and one fix spreads to that hand.
- The Describer drafts each folder to DACS, citing its page.
- Folder 4 names three students on probation; the Sensitivity agent holds it, citing FERPA.
- Tomas confirms; the head of special collections approves the restriction.
- Eleven folders go live with accessible transcripts; folder 4 stays restricted.
- Every line keeps its approver and source page, ready to export.
- The aim: four times more linear feet per archivist hour.
- Built on Botminds: agents, data, screens and governance on one platform.
The narration as written for the film. Captions are on the player.