- Support
- Integrations
- Document Understanding
Document Understanding
Oracle Cloud integration · 12 node(s).
00Overview
Pull the content out of documents straight from a flow — extract text, tables and form key-value pairs, classify a document's type or language, and detect its structural elements. Analyse a single file inline for an immediate result, or start an asynchronous processor job that reads a document from Object Storage and writes the results back to a bucket. You can also manage the projects and custom models behind Document Understanding: create, list, fetch, update, move between compartments and delete projects, and look up model details and training state.
Every field below is exactly what you see in the Flomation editor. Fields marked ● live picker let you choose from a list pulled live from your account — no IDs to look up.
01Connecting Document Understanding
- Every action begins with an Authentication choice: Connect Oracle Cloud (recommended) uses a managed connection you set up once, while API signing key (advanced) lets you supply the raw OCI signing-key fields directly on the node.
- For Connect Oracle Cloud, add an Oracle Cloud connection in Flomation's environment settings and follow its guided setup once to authorise Flomation in your tenancy — you provide your tenancy and region, and the connection handles the user and signing key for you.
- After setting it up, run the connection's test check — a brand-new signing key can take a short while to propagate, so retry the test once if it does not pass on the first attempt.
- In each Document Understanding node, leave Authentication on Connect Oracle Cloud and pick your connection in the Oracle Cloud connection field — the tenancy, user, region, fingerprint and signing key are filled in automatically.
- Set the Compartment OCID on every node to the compartment that holds (or will hold) your projects, models and jobs — it scopes both what you can see and where new resources are created.
- To use a raw key instead, switch Authentication to API signing key (advanced) and take Tenancy OCID, User OCID, Region and Key Fingerprint from Identity & Security → Users → your user → API Keys in the OCI Console; store the PEM in a Flomation environment secret and select it in the node's Private Key (PEM) field, adding a Private Key Passphrase only if the key is encrypted.
| Field | Type | Details | |
|---|---|---|---|
| Authentication | string | Connect Oracle Cloud, API signing key (advanced) | |
| Oracle Cloud connection | credential | Pick a connected Oracle Cloud account | |
| Region | string | e.g. uk-london-1 | |
| Private Key (PEM) | secret | The API signing private key — full PEM, incl. BEGIN/END lines | |
| Private Key Passphrase | secret | Only if the key is encrypted (optional) | |
| Tenancy OCID | string | ocid1.tenancy.oc1..aaaa… | |
| User OCID | string | ocid1.user.oc1..aaaa… | |
| Key Fingerprint | string | aa:bb:cc:… fingerprint of the uploaded API key |
Pick an Environment on your flow (Flow Settings → Environment) so the secret resolves. Secret fields never show the value — they reference ${secrets.your_secret}.
02Analyze
OCI Document Understanding: Analyze Document
oracle/documentunderstanding/analyze_document · Action
Synchronously analyze an inline (base64) document — extract text, tables and form key-values, and/or classify the document type — returning the raw analysis result.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Document (Base64) | text | Required | The document bytes, Base64-encoded (PDF, TIFF, PNG or JPEG). A data: URI prefix is accepted. |
| Feature Types | string | Required | Comma-separated: TEXT_EXTRACTION, TABLE_EXTRACTION, KEY_VALUE_EXTRACTION, DOCUMENT_CLASSIFICATION |
Returns: tool_result, result, success, error
03Model
OCI Document Understanding: Get Model
oracle/documentunderstanding/model_get · Action
Fetch a single Document Understanding custom model by its OCID — its type, version, training details and lifecycle state.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Model OCID | string | Required | ocid1.aidocumentmodel.oc1..aaaa… |
Returns: tool_result, model, id, lifecycle_state, success, error
OCI Document Understanding: List Models
oracle/documentunderstanding/model_list · Action
List the custom Document Understanding models in a compartment. Optionally scope to a project and filter by exact display name or lifecycle state. Walks pagination up to a safe cap.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… (use the tenancy OCID for the root) |
| Project OCID Filter | string | Only models in this project (optional) | |
| Display Name Filter | string | Only models with this exact name (optional) | |
| Lifecycle State | string | Filter by state (optional) — choices: Creating, Updating, Active, Deleting, Deleted, Failed | |
| Page Size | string | Max items per page (optional) |
Returns: tool_result, models, count, truncated, success, error
04Processor
OCI Document Understanding: Cancel Processor Job
oracle/documentunderstanding/processor_job_cancel · Action
Cancel an in-progress Document Understanding processor job by its OCID, stopping further processing.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Processor Job OCID | string | Required | ocid1.aidocumentprocessorjob.oc1..aaaa… of the job to cancel |
Returns: tool_result, id, success, error
OCI Document Understanding: Create Processor Job
oracle/documentunderstanding/processor_job_create · Action
Start an asynchronous processor job that analyses a document in Object Storage (text, table, key-value, classification, language or elements) and writes results to an output prefix. Returns the job in an ACCEPTED state — poll Get Processor Job until SUCCEEDED.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Input Namespace | string | Required | Object Storage namespace of the input document |
| Input Bucket | string | Required | Bucket holding the input document |
| Input Object | string | Required | Object (file) name of the document to analyse |
| Output Namespace | string | Required | Object Storage namespace for the results |
| Output Bucket | string | Required | Bucket to write results to |
| Output Prefix | string | Required | Folder/prefix under which results are written |
| Feature Types (CSV) | string | Required | TEXT_EXTRACTION,TABLE_EXTRACTION,KEY_VALUE_EXTRACTION,DOCUMENT_CLASSIFICATION,LANGUAGE_CLASSIFICATION,DOCUMENT_ELEMENTS_EXTRACTION |
Returns: tool_result, processor_job, id, lifecycle_state, success, error
OCI Document Understanding: Get Processor Job
oracle/documentunderstanding/processor_job_get · Action
Fetch a single Document Understanding processor job by its OCID — its lifecycle state and processing progress.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Processor Job OCID | string | Required | ocid1.aidocumentprocessorjob.oc1..aaaa… |
Returns: tool_result, processor_job, id, lifecycle_state, percent_complete, success, error
05Project
OCI Document Understanding: Change Project Compartment
oracle/documentunderstanding/project_change_compartment · Action
Move a Document Understanding project into a different compartment — the project keeps its OCID, only its compartment placement changes.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Project OCID | string | Required | ocid1.aidocumentproject.oc1..aaaa… (the project to move) |
| Destination Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… (where to move the project) |
Returns: tool_result, id, destination_compartment_id, success, error
OCI Document Understanding: Create Project
oracle/documentunderstanding/project_create · Action
Create a Document Understanding project to hold custom models. Returns the project in a CREATING state plus a work-request id — poll Get Project until ACTIVE.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Display Name | string | A name for the project (optional) | |
| Description | string | An optional description of the project |
Returns: tool_result, project, id, lifecycle_state, work_request_id, success, error
OCI Document Understanding: Delete Project
oracle/documentunderstanding/project_delete · Action
Delete a Document Understanding project by its OCID — returns the work-request tracking the teardown.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Project OCID | string | Required | ocid1.aidocumentproject.oc1..aaaa… of the project to delete |
Returns: tool_result, id, work_request_id, success, error
OCI Document Understanding: Get Project
oracle/documentunderstanding/project_get · Action
Fetch a single Document Understanding project by its OCID — its display name, description, lifecycle state and timestamps.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Project OCID | string | Required | ocid1.aidocumentproject.oc1..aaaa… |
Returns: tool_result, project, id, lifecycle_state, success, error
OCI Document Understanding: List Projects
oracle/documentunderstanding/project_list · Action
List the Document Understanding projects in a compartment. Optionally filter by exact display name or lifecycle state. Walks pagination up to a safe cap.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… (use the tenancy OCID for the root) |
| Display Name Filter | string | Only projects with this exact name (optional) | |
| Lifecycle State | string | Filter by state (optional) — choices: Creating, Updating, Active, Deleting, Deleted, Failed |
Returns: tool_result, projects, count, truncated, success, error
OCI Document Understanding: Update Project
oracle/documentunderstanding/project_update · Action
Partially update a Document Understanding project — change only the display name or description you supply; blank fields are left unchanged. Asynchronous: returns a work-request id, poll Get Project until it settles.
| Field | Type | Details | |
|---|---|---|---|
| Compartment OCID | string | Required | ocid1.compartment.oc1..aaaa… |
| Project OCID | string | Required | ocid1.aidocumentproject.oc1..aaaa… — the project to update |
| Display Name | string | New name (leave blank to keep unchanged) | |
| Description | string | New description (leave blank to keep unchanged) |
Returns: tool_result, id, work_request_id, success, error
06Notes & Limitations
Behaviours and constraints worth knowing before you build with these nodes.
- Create Processor Job returns immediately in an
ACCEPTEDstate, so the flow must poll Get Processor Job until its lifecycle state reachesSUCCEEDEDbefore reading the results written to the output prefix. - Analyze Document runs synchronously on a single inline document and covers text, table, key-value and classification, whereas Create Processor Job runs against Object Storage and additionally supports language classification and document-element extraction.
- Project create, update and delete are asynchronous and return before the change settles, so follow them with Get Project to confirm the project has reached
ACTIVE(or been fully deleted) before dependent steps run. - List actions walk pagination only up to a fixed page cap, so results in very large compartments can be incomplete unless you narrow them with the display-name or lifecycle-state filters.