➕ Document Attributes¶

Document attributes describe the identity, classification, lifecycle, and context of a document as it moves through your AIForged flows. Use them to route work, enforce policy, and correlate with external systems.

Tip

Attributes are different from Parameter values. Attributes describe the document and its lifecycle; Parameters are field values (outputs) extracted by services (e.g., invoiceNumber, total).

Quick reference¶

Field	Description	Example
Id	Unique identifier for the document (system‑generated).	1234
Master ID	Links related documents into a hierarchical or cross‑source flow. Useful for grouping emails and their extracted attachments, or multi‑step pipelines.	1234
Filename	The stored filename.	invoice_12345.pdf
File Type	File extension.	pdf
Content Type	MIME type of the binary.	application/pdf
Category	Classification assigned to the document (by a classifier or manually).	Invoice, ID, Payslip
Type	Document type used by downstream extractors/definitions.	Invoice (Contoso), Bank Statement
Usage	Indicates where the document is used within the flow. Common examples include inbox/outbox/training/result.	Inbox, Outbox, Training, Result
Status	The current lifecycle state of the document.	Created, Processing, Completed, Error
Comment	Free‑text notes for operators and audit context.	Duplicate detected via checksum
Result	Output payload or summary set by services (often OCR/Classification). May reference generated files or JSON.	JSON with extracted fields
External ID	Optional, user‑defined identifier for correlating with third‑party systems.	ERP‑INV‑000123
Date Created	When the document was added to AIForged.	2025‑08‑12 10:14 UTC
Date Modified	Last time the document or its data changed (processing, verification, etc.).	2025‑08‑12 10:21 UTC

Info

Exact Status and Usage values can vary by solution. Use them consistently for routing and reporting. If you need a canonical list for your tenant, we can add a reference table.

Where these attributes come from¶

Id, Filename, File Type, Content Type
- Set on upload/ingest (manual upload or Scrapers like Email/OneDrive).
Category, Type
- Set by Classification or Extract & Verify services (or manually edited when permitted).
Usage
- Set/updated by your pipeline as a document moves (e.g., Inbox → Outbox).
Status
- Managed by the platform and services during processing (e.g., Created → Processing → Completed/Error).
Result
- Set by services that produce outputs (OCR, Classification, Extract & Verify, LLM Extractor via post‑processing).
External ID
- Set by you (upload, API, SDKs) to map documents to external systems.
Master ID
- Assigned/propagated by flows to group related items (e.g., email parent to child attachments).

Tip

Use External ID to keep a stable link to your ERP/CRM. Use Master ID to group multi‑file work (e.g., email + attachments, split pages, or downstream derivatives).

Relationships and hierarchy¶

Parent/child documents
- Email “parent” → Attachment “children” (created by the Attachment utility).
Master ID grouping
- Keeps disparate but related documents linked across services and locations.
Derived artifacts
- Utilities (Digitizer, PDF Converter) may create transformed versions while preserving identity and traceability.

Viewing and editing in Studio¶

Document page header
- Shows Id, Filename, Type/Category, Status, Usage, and timestamps.
Editable fields
- Comment and External ID (subject to roles/permissions).
Operational context
- You’ll often see Parameter values (extracted data) alongside attributes for verification.

Tip

If attributes or actions appear disabled, confirm you’re in the correct Tenant and that your role (Owner, Administrator, Developer) allows edits.

Programmatic access (REST, SDKs, RPA)¶

AIForged is built on REST APIs and also offers SDKs for .NET and JavaScript/TypeScript, plus UiPath and Power Automate integrations.

Typical operations
- Get a document by Id, update Comment/External ID, query by Status/Usage/Category, fetch Results and Parameter values.
REST pattern (generic)
- GET/PUT https://{your-endpoint}/api/{documents}/{id}
SDKs
- .NET (C#) and JavaScript/TypeScript provide strongly‑typed access to documents, datasets, services, and work items.

Tip

Prefer SDKs for type safety and faster development. Use REST for lightweight scripting or custom integrations. RPA (UiPath/Power Automate) can read/update attributes inside broader enterprise flows.

Documents can have multiple data blobs linked to them (originals, processed outputs, training data, etc.). These are typed via DocumentDataType:

Name	Value	Description
Image	0	Original document images/data blob/sound file.
Result	1	Result data produced by processing.
Page	2	Per‑page data associated with the image.
Backup	3	Backup of an image.
Definition	10	Definition/configuration data.
Training	11	Training artifacts or examples.

Use these types to retrieve the right artifact (e.g., final JSON, searchable PDF) for your exports or audits.

Best practices¶

Keep Category and Type clean and consistent—they drive routing, model selection, and reporting.
Use Usage and Status to orchestrate flows and dashboards (e.g., “Processing backlog,” “Ready for QA”).
Populate External ID on ingest to simplify reconciliation with external systems.
Leverage Master ID to hold multi‑file jobs together across services.
Add concise Comments for human context—this helps during HITL and audits.