integrations · one.semurg.io/integrations

Everything in. One store.

Building, connectors in progress

Semurg Integrations is the surface that brings the rest of your estate in: databases, files, APIs, message streams and SaaS systems, landing as the same uniform 64-byte containers as everything else. There is no staging warehouse and no second copy to keep in sync, because what arrives IS the store. The same data is then readable through every face at once, so a record ingested from a relational source is immediately a graph node, a search hit and an analytics row without a second load.

What you will build here

Two-way, not just importData comes in and goes back out through the same connector, so Semurg can sit beside a system of record rather than demanding you replace it first.

Connectors as extensionsA connector is a Tier-3 extension on the one substrate, inward-only, owning no store of its own, so adding one can never fork the data model.

Land once, read every wayAn ingested record is queryable through graph, SQL, key-value, search, vector and analytics immediately, with no per-face re-index and no export step.

Honest status

✓Done, the universal ingest path and the one data model every connector lands into, plus round-trip proven byte-exact against several external databases.
●In progress, the connector registry and the two-way sync surface.
○Queued, the public connector catalogue, scheduling and per-connector throughput figures.

No connector count and no throughput figure ships here until each one is measured on our reference box and reproducible on yours, the same rule every other Semurg number follows. Until then this page says what is built and what is not, and nothing more.

The ingest path underneath is already real and reproducible today: run the board yourself, or bring your own data in through the HTTP API.