Extract event

Delivered when an extract function completes processing of a document and produces structured JSON. The body's transformedContent matches the function's outputSchema.

For signing-secret setup and retry semantics that apply to every variant, see the Webhooks overview.

Request Body

application/json

TypeScript Definitions

Use the request body type in TypeScript.

functionCallTryNumber?integer

The attempt number of the function call that created this event. 1 indexed.

eventID*string

Unique ID generated by bem to identify the event.

createdAt?string

Timestamp indicating when the event was created.

Formatdate-time

referenceID*string

The unique ID you use internally to refer to this data point, propagated from the original function input.

inboundEmail?

The inbound email that triggered this event.

metadata?

eventType?string

Value in"extract"

functionCallID?string

Unique identifier of function call that this event is associated with.

functionID*string

Unique identifier of function that this event is associated with.

functionName*string

Unique name of function that this event is associated with.

functionVersionNum?integer

Version number of function that this event is associated with.

callID?string

Unique identifier of workflow call that this event is associated with.

workflowID?string

Unique identifier of workflow that this event is associated with.

workflowName?string

Name of workflow that this event is associated with.

workflowVersionNum?integer

Version number of workflow that this event is associated with.

inputType?string

The input type of the content you're sending for transformation.

Value in

"csv" | "docx" | "email" | "heic" | "html" | "jpeg" | "json" | "heif" | "m4a" | "mp3" | "pdf" | "png" | "text" | "wav" | "webp" | "xls" | "xlsx" | "xml"

transformationID?string

Unique ID for each transformation output generated by bem following Segment's KSUID conventions.

s3URL?|

Presigned S3 URL for the input content uploaded to S3.

inputs?array<>|

Array of transformation inputs with their types and S3 URLs.

transformedContent*object

The transformed content of the input. The structure of this object is defined by the function configuration.

correctedContent?||array<>|||||

Corrected feedback provided for fine-tuning purposes.

invalidProperties?array<>

List of properties that were invalid in the input.

itemOffset*integer

The offset of the first item that was transformed. Used for batch transformations to indicate which item in the batch this event corresponds to.

itemCount*integer

The number of items that were transformed. Used for batch transformations to indicate how many items were transformed.

fieldBoundingBoxes?object

Per-field bounding boxes. A JSON object mapping RFC 6901 JSON Pointer paths (e.g. "/invoiceNumber", "/items/0/price") to the document regions from which each extracted value was sourced.

fieldConfidences?object

Per-field confidence scores. A JSON object mapping RFC 6901 JSON Pointer paths (e.g. "/invoiceNumber") to float values in the range [0, 1] indicating the model's confidence in each extracted field value.

avgConfidence?|

Average confidence score across all extracted fields, in the range [0, 1].

Response Body

Example Requests

V3 event variants that do not exist in the shared Event union.

ExtractEvent and ClassifyEvent are emitted only by V3-era function types (extract and classify). The shared Event union in specs/events/models.tsp predates these types and continues to describe V2 / V1-alpha responses verbatim; V3 response payloads add the new variants via the EventV3 union below while keeping every shared variant intact for backward compatibility.

POST/extract

Extract event

Request Body

Response Body

2XX

Request Bodyapplication/json

See also