Infrastructure

Arvest is a web app backed by an open API, run on public research infrastructure. Documents come in from libraries, from the web and from your own computer; everything is described in IIIF, the open standard for cultural heritage, so it can go out again to any IIIF tool.

  1. 1Documents from everywhere
  2. 2The app, in your browser
  3. 3An open API
  4. 4Kept at Huma-Num
  5. 5Shared with the world
1

Documents from everywhere

Bring in IIIF manifests from libraries, archives and museums, add any image, video or sound by its web address, or upload your own files from your computer. A linked document stays where it is: Arvest reads it, it does not copy it.

2

The app, in your browser

Nothing to install: an account with an email address opens the whole workspace, where documents are placed, compared, annotated and linked. The first version of Arvest is built on our fork of the IIIF viewer Mirador, which plays audio and video, with plugins for annotating images and audiovisual content. Arvest v2, in beta, is a new workspace built on the same API.

3

An open API

The app and your scripts speak to the same REST API, written in NestJS: accounts, groups, projects, media, manifests, metadata and annotations. It is documented with Swagger, and a Python package makes it easy to use in scripts and machine learning notebooks.

4

Kept at Huma-Num

Arvest runs on a virtual machine at Huma-Num, the French state-funded infrastructure for the digital humanities. It keeps the database of accounts and projects, the files you upload and the IIIF manifests that describe them, and a PeerTube instance for large videos.

5

Shared with the world

Documents, manifests and whole projects can be shared by link, with permissions that keep you in control of who sees and who edits. What you share is IIIF, so it also opens in other IIIF viewers and tools.