Unlock audiovisual collections and advance user discovery with AI-generated captions and transcripts in Specto
Libraries can hold years of valuable audio and video content that remains difficult to find or use.
Recorded interviews, lectures, and other audiovisual materials often contain rich information, but limited metadata can make that information hard to discover. Creating captions, transcripts and detailed descriptions manually is time and resource intensive for library staff.
We’ve introduced new AI-powered capabilities in Specto that help libraries bring this content to life. Staff can generate captions and searchable transcripts for audio and video content, at scale, use transcript content to enrich metadata, while the transcript text is indexed in and discoverable through Primo.
Generate captions and searchable transcripts
Library staff can select audio or video titles in Specto and run the AI Generate Captions feature. Specto then generates captions and transcripts and adds them to the selected titles.
During playback, captions appear over the recording, while a separate transcript pane lets users read and search the spoken content.
This can help users follow recorded content more easily while reducing the staff time needed to create captions and transcripts manually.

Use transcripts to support metadata enrichment
Specto uses AI to identify entities mentioned in the transcript and suggest links to authority records. Library staff review and approve the recommendations, adding their own knowledge where needed. They can then run a complementary AI-assisted process to further enrich the metadata record, which the cataloger also reviews and approves.
Imagine a library with a large collection of interviews. Describing each interview in detail would require staff to listen to lengthy recordings and identify the subjects discussed. AI-generated transcripts and extracted entities reduce the need for manual transcription and allow staff to focus on their cataloging work.

Bring more context from recordings into Primo discovery
Transcript content is indexed in Primo, making the spoken content searchable and giving students and researchers another way to find relevant audio and video material. Enriched metadata can also help users find recordings when they search for the people, organizations and places mentioned in them.
After opening a resource, users can view and search the transcript alongside the recording and move through the content more easily. Related entities can provide additional context and lead users to other materials.
Make more of existing audiovisual collections
For libraries with new, large or older audio and video collections, the challenge is often not a lack of valuable material. It is the work required to describe that material well enough for students and researchers to find and use it.
AI-generated captions and transcripts in Specto can reduce manual effort, support accessibility and make information inside recordings easier to discover through Primo.
Institutions remain responsible for determining which materials from their collections they choose to process using these capabilities and for managing access to those materials in accordance with their policies and applicable requirements.
Want to see how Specto helps libraries bring their digital collections to life?
Watch the demo to discover how easy it can be to manage, showcase, and connect audiences with your digital collections.