Skip to main content
DataErai organizes research data in a clear hierarchy. Understanding these few concepts makes every other guide easier to follow.

Organization

Your institution or lab’s home. Holds projects, members, and storage.

Project

A workspace for a study, grant, or effort.

Collection

A nestable container that groups assets within a project.

Asset

A file (or bundle) plus its metadata, notes, and relationships.

Dataset

A versioned, analysis-ready view of your data.

Transfer

An upload or download of file content.

The hierarchy

Data nests from broad to specific: Organization → Project → Collection → Asset → Content An organization contains projects; a project contains collections (which can nest inside one another); collections hold assets; and an asset holds its file content. Access granted higher up cascades down, so a person who can read a project can read the collections and assets inside it — unless you override access on a specific item.

Organizations and members

An organization is the top-level home for your institution, lab, or company. Every member of an organization has a role:
RoleWhat it allows
MemberWork within the organization’s projects according to the permissions they’re given
AdminManage the organization and its members
See Organizations & members.

Projects

A project is the primary workspace — typically one per study, grant, or effort. Within a project, members have a role:
RoleWhat it allows
MemberView and work with project data they have access to
ManagerCreate and manage content in the project
AdminFull control of the project, including sharing and membership
The person who creates a project is its Admin. See Projects & members.

Collections

A collection is a folder-like container that groups assets inside a project. Collections can nest, so you can mirror how your team already thinks about its data — by experiment, instrument, sample, or date. See Collections.

Assets and content

An asset is the core unit of data. It combines:
  • Metadata — title, description, tags, and structured fields (often extracted automatically from instrument files).
  • Content — the actual file or files. An asset can exist before its content is uploaded, which is useful for metadata-first workflows.
  • Notes — a Markdown note or README describing protocols, settings, and caveats.
  • Relationships — links to other assets that capture provenance (which data produced which).
See Assets & content and Provenance.

Datasets and versions

A dataset is a curated, versioned view of data prepared for analysis. Each new version is an immutable snapshot, and DataErai can analyze a version to infer its structure (for example, the columns and types in a table) so it’s ready to explore and visualize. See Datasets.

Transfers

A transfer is the movement of file content — an upload or a download. DataErai is built for large scientific files, so transfers are high-throughput and resumable: if a transfer is interrupted, it picks up where it left off. You can transfer in the browser, or with the desktop app and CLI.

Storage: repositories and allocations

Your data physically lives in a repository, and your team is given a storage allocation (how much space you can use) within it. Most of the time this is set up for you; you’ll mainly notice it when choosing where new data goes or checking how much space you’ve used. See Storage.

Groups and permissions

Access is controlled per object — project, collection, asset, or dataset — using three roles:
RoleWhat it allows
ReadView metadata, content, and notes
WriteRead, plus edit metadata, content, and notes
AdminWrite, plus share the object with others
You can grant a role to an individual, to a group of people, or to the public (public access is always read-only). Datasets can also mark sensitive restricted fields that require an extra grant to view. See Sharing & permissions.

Attribution

DataErai connects your data to the research record. Link assets and projects to the grants that funded them and the publications (works) they support, and credit the people who contributed. See Attribution.