Knowledge that lasts.
Underlay is a public registry where organizations publish versioned snapshots of their structured data — making it permanently discoverable, verifiable, and citable. Research datasets, publication archives, open knowledge: published once, preserved indefinitely.
Permanent
Every version of your data is stored immutably. Research doesn't disappear when a project ends or a server goes down. Once published, it's preserved.
Verifiable
Content-addressed storage means anyone can confirm that data hasn't been altered. Cryptographic hashes at every level make tampering evident and trust auditable.
Discoverable
Structured data with shared schemas across collections. Browse, search, and export — every dataset is openly accessible through both the web and a REST API.
Built for
Research institutions
Preserve datasets beyond the life of a grant. Publish versioned snapshots that are citable and independently verifiable.
Academic publishers
Archive publication metadata, review data, and supplementary materials in a format that's structured, searchable, and open.
Open data organizations
Share curated datasets with the public. Underlay handles versioning, integrity, and access — so you can focus on the data itself.
Developers
Build on open knowledge. Pull snapshots via API, integrate with existing workflows, or run your own Underlay instance.
How it works
Publish
Push structured data via a simple REST API. Define your schema, add records and files, and create a versioned snapshot.
Preserve
Each version is validated, deduplicated, and stored immutably. Files are content-addressed. Every byte is accounted for.
Discover
Anyone can browse collections, inspect schemas, view diffs between versions, and export full archives. The data is the interface.
Open source
MIT licensed. Run your own instance, contribute, or push data to the canonical host at underlay.org.
Built by Knowledge Futures
A 501(c)(3) nonprofit building open infrastructure for the production, curation, and preservation of knowledge.