Library and Archives Canada / Bibliothèque et Archives Canada

Pilot project to categorize a sample of 25,000 digital archival files

Proof of Concept with Oproma to generate metadata for a sample of 1000 digitized government publications.

Improve access to Department of Indian Affairs and Northern Development historical records (RG10) via transcription of ~6 million pages, relevant to Indigenous history.

Automatically classify ~10 million documents from the former LAC Network Shared Drives (NSD) and eliminate ROT (Redundant, Obsolete, Trivial) information.

Create a full index of the 1931 Canadian Census (~10 million names) using handwriting recognition technology and AI for transcription.

Pilot transcription projects using Transkribus AI model to improve access to historical records: (1) Index of Canadian Newspapers and (2) Order-in-Council Registers (1925 to 1940).

Generative AI chatbot for unclassified government work, trained on Canadian data (including 50 TB from LAC Government of Canada Web Archive) to enhance employee productivity, sovereignty, and digital independence.