Skip to main content
A Beam pipeline config describes the full data flow: where data comes from (source), how it’s processed (transforms), and where it goes (sinks).

BeamConfig

The top-level object returned by all config endpoints.

CreateBeamConfigRequest

Used as the request body for Create pipeline. Same as BeamConfig but without auto-generated fields.

UpdateBeamConfigRequest

Used as the request body for Update pipeline. The shape is identical to CreateBeamConfigRequest — only the editable fields. The config_id is taken from the URL path, not the body. Server-managed fields (id, owner_org_team_user_id, organization_id, created_at, updated_at) are preserved from the existing record and ignored if present in the body.

PipelineConfig


Source

Connects to Allium’s Datastreams. Select a blockchain and entity type.

Supported chains and entities

More chains and entities are being added regularly. Check the Datastreams catalog for the latest availability, or use the GET /api/v1/beam/sources endpoint (requires a Beam-enabled organization and API key).

Transforms

Transforms process data in order. Two types are available. Each transform has an auto-generated uid — omit it when creating new transforms and the server will assign one.
Filters data by matching a field value against a set. Only records whose extracted value exists in your set pass through. Sets support 10M+ values.
Common filter_expr patterns:Filter values are managed separately via the filter values endpoints. Changes take effect immediately — no redeploy needed.
Use lowercase values when filtering by addresses, labels, or symbols.

Sinks

Sinks define where processed data is delivered. A pipeline must have exactly one sink — the sinks array accepts a single entry. Each sink has an auto-generated uid — omit it when creating new sinks. Five types are available:
Delivers to an Allium-managed Kafka topic. After deployment, you receive connection credentials and consumer code snippets.
Contact support@allium.so if you need a specific sink type.

ID fields and creation behavior

When creating resources (pipelines, transforms, sinks), omit the id / uid field and the server will auto-generate one. When updating, include the id / uid to reference the existing resource.