A
Interview workspace
Guided capture, release context, narrator preferences, and collection fields in one place.
Oral history · responsibly assisted
Keep living
memory alive.
Tavorelia helps cultural memory teams record, govern, preserve, and publish adult oral histories—with consent and provenance built in.
Pre-launch Building with museums, archives, libraries, and cultural nonprofits.
The quiet risk
A recording without durable context, consent, and access rules is not yet a trustworthy public memory.
Why oral traditions matter ↗From spoken memory to governed collection.
Designed around the work curators already do—not around a black-box generation button.
- 01
Record
Capture a rights-cleared adult interview, its context, and the narrator’s choices in one guided session.
- 02
Transcribe
Create a time-aligned draft transcript, surface uncertain passages, and keep every edit reviewable.
- 03
Govern
Track consent, provenance, access rules, permitted uses, and withdrawal requests beside the source.
- 04
Publish
Shape approved material into searchable collections, listening pages, story maps, and exhibition drafts.
Less production friction.
More curatorial control.
B
Reviewable AI drafts
Time-aligned transcription, translation, summaries, and metadata with confidence cues and human approval.
C
Rights-aware archive
Consent versions, provenance, access levels, embargoes, permitted reuse, and withdrawal history.
D
Public story formats
Approved listening pages, thematic collections, exhibit drafts, transcripts, and accessible audio guides.
AI can assist the archive. It must not overrule it.
Every automated output is a proposal. Every publication decision stays with people who hold responsibility for the collection and the narrator.
- Narrator agency
- People decide how their testimony may be used, shared, corrected, or withdrawn.
- Source before summary
- AI-generated metadata and summaries remain traceable drafts until a human curator approves them.
- No training by default
- Institutional collections are not used to train shared models by default.
- Synthetic stays separate
- Generated voices and invented testimony do not belong inside the historical record.
Useful automation, clearly bounded.
Planned for the pilot
Speech-to-text, translation drafts, semantic search, metadata suggestions, accessibility support, and curator-assisted exhibit copy.
Outside the pilot
No voice cloning, no invented testimony, no automatic publication, and no training shared models on institutional archives by default.
Proposed pilot pricing
A small-team starting point.
Pricing is a working proposal and will be shaped with pilot partners before launch.
Institution starter
$149 / month
- 3 curator seats
- Up to 25 interview hours / month
- Consent and provenance records
- One public collection
- Exportable media and metadata
Proposed, not yet for sale. Larger and community-sponsored programs will be scoped separately.
Pre-launch, with the hard questions first.
Now
Discovery
Interviewing memory institutions and documenting governance requirements.Next
Private pilot
Record-to-review workflow with a small group of adult oral-history projects.Then
Governed publishing
Rights-aware collections, accessible listening pages, and export tools.No customers, revenue, or production deployments yet. The waitlist is for institutions that want to help shape the pilot.
Before a memory enters the system.
Who is Tavorelia for?
Museums, archives, libraries, universities, cultural nonprofits, diaspora associations, and community-history teams working with adult narrators.
Does AI decide what gets published?
No. AI outputs are drafts. A responsible human reviewer approves edits, metadata, access rules, and publication.
Will you clone narrators’ voices?
No. Voice cloning is outside the pilot. Synthetic material must never be confused with an authentic historical record.
What happens if a narrator changes their mind?
The planned workflow keeps consent versions and withdrawal requests connected to every affected item, so curators can act consistently and document the decision.
Which languages will be supported?
Multilingual transcription and translation are planned, but the initial language set will be chosen with pilot partners and validated by human reviewers.
Help build a more accountable home for living memory.
Tell us about your institution and collection. We’ll contact a small set of teams for discovery conversations and early pilot planning.
Prefer email? ceo@tavorelia.online