Get customers
Answer customers
Work for you
Reach people
Give agents tools
Run and ship
Automate document extraction, agent runs and human review tasks
List the connected account's agents.
2 fields
Read an agent and its resource IDs.
1 field
List one page of runs for an agent.
4 fields
Start a run, optionally passing variables through to its output.
Read a run's status and output file references.
Archive a run using the documented status patch.
Delete one explicitly selected agent run.
Read an agent's processing statistics.
List one page of documents, optionally within a dataset.
5 fields
Create a document record with optional inline base64 file content. If content is omitted, upload separately to the returned fileUrl before expecting predictions. Creating the record alone does not upload a file.
8 fields
Read a document's metadata and file references.
Patch only supplied document metadata fields.
6 fields
Delete one explicitly selected document.
List stored predictions, optionally filtered by model.
Extract structured data from an existing document using a model.
Read a stored prediction and its extracted values.
List the account's extraction models.
Read an extraction model's configuration.
List the account's human review configurations.
Read a human review configuration.
List one page of review tasks within a validation.
Create a review task; reuse an existing task when one already exists.
Read a task and its current review state.
Update review output or status, including completing or cancelling a task.
7 fields
