Our sources

We use NASA.
They use Shutterstock.

Every clip in every Verika production is sourced from a named institutional archive — with an accession number, a licence record, and a verifiable chain of custody. No stock footage libraries. No unnamed sources. No synthetic media.

30+
Institutional archives integrated
Public domain · Creative Commons · Government open data
Named — every archive is cited by name
Logged — accession number on every clip
Cleared — licence status documented
Signed — C2PA manifest on every delivery

…and many more institutional, government, and scientific archives — integrated directly into the engine, with new sources added continuously.

Each source has a dedicated adapter handling API authentication, search, pagination, and licence inference.

How it works

From narration to named clip.

01

Script drives the search.

Each sentence of narration is analyzed semantically. The engine identifies what needs to be shown — a location, a person, an event, a concept — and determines which archives are most likely to have it.

02

Archives are queried in parallel.

30+ source adapters search simultaneously. Each adapter handles its archive's specific API, authentication, pagination, and rate limits. Results are scored for visual relevance using semantic matching.

03

Every clip is logged and signed.

The winning clip is downloaded, its source recorded with accession number and licence, and it's woven into the timeline. At render, the full provenance chain is signed with a C2PA manifest.

See it in action

Real footage.
Real proof.

Request a free pilot and see what a Verika production looks like — every clip from a named archive, every frame signed.

Request your free pilot →