Ingestion

Document ingestion endpoints. TamaDB uses upsert semantics — there is no separate create vs update. If the document exists, incoming fields are merged.

Quick Reference

Method	Path	Summary
POST	/api/v1/ingest/{documentType}	Direct ingestion via Document_Type
POST	/api/v1/ingest/{documentType}/{channelId}	Channel ingestion (with enrichment)

POST

/api/v1/ingest/{documentType}

Ingest a document directly via its Document_Type. Uses upsert semantics — creates the document if new, merges fields if it already exists.

Request Body

Field	Type	Required	Description
document	object	Yes	The document payload to ingest
versionNo	number	No	Expected version for optimistic concurrency control
ttl	number \| "infinite"	No	Per-document TTL override (seconds)

Response — New Document

201 Created

{
  "dId": "LIVING-ROOM-01",
  "versionNo": 1,
  "created": true
}

Response — Updated Document

200 OK

{
  "dId": "LIVING-ROOM-01",
  "versionNo": 2,
  "created": false
}

Errors

400 Bad Request 404 Document_Type not found 409 Version conflict 422 Schema validation failed

Note: D-Id is extracted from the document using the schema's identityKind annotations. For composed D-Ids (parent-key + primary-key), the response returns an array.

POST

/api/v1/ingest/{documentType}/{channelId}

Ingest a document via a specific channel. Enrichment bindings are applied to the payload before validation.