File Intelligence

What it does

File Intelligence is where you set how fileAI processes documents in a project: how files are split, which language to prioritise, which AI parsing model reads them, and how they are classified. These settings are the project’s defaults for every file.

When to use it

You need fileAI to return values in one specific language for every file in the project, instead of As-is.
You need a different AI parsing model as the project default, such as Beethoven_Direct_Form_Filling (GP2.5) for specialised form workflows.
You receive files that each hold several documents, or workbooks with one record per worksheet, and want each processed as its own document.

New projects start with recommended defaults, so you only need to change them for specific requirements.

How it works

Open File Intelligence

Go to AI Drive in the fileAI app, then select File Intelligence in the left navigation.

Change the defaults

Edit General Settings in the app, or update the AI parsing model, target language and file splitting with PATCH /v1/project/setting.

Process files

The defaults apply to every file in the project. Four of them can be overridden for a single upload: ocrModel, schemaLocking, isSplit and isSplitExcel. See Upload options.

File Intelligence settings apply to the whole project. You cannot set them per folder or per file here.

Try it

PATCH /v1/project/setting updates three of the defaults: ocrModel, formTargetLanguage and isSplit. Send ocrModel and formTargetLanguage together; a request missing either returns 422. Both take the names the API lists, such as Beethoven_ENG_GP25.2 and English or As-is.

curl -X PATCH "$FILEAI_BASE_URL/project/setting" \
  -H "x-api-key: $FILEAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "ocrModel": "Beethoven_ENG_GP25.2", "formTargetLanguage": "English" }'

import os, requests

BASE = "https://api.orion.file.ai/prod/v1"

r = requests.patch(
    f"{BASE}/project/setting",
    headers={"x-api-key": os.environ["FILEAI_API_KEY"]},
    json={"ocrModel": "Beethoven_ENG_GP25.2", "formTargetLanguage": "English"},
)
result = r.json()

const BASE = "https://api.orion.file.ai/prod/v1";

const result = await fetch(`${BASE}/project/setting`, {
  method: "PATCH",
  headers: {
    "x-api-key": process.env.FILEAI_API_KEY,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ ocrModel: "Beethoven_ENG_GP25.2", formTargetLanguage: "English" }),
}).then((r) => r.json());

Set FILEAI_BASE_URL (or BASE) to https://api.orion.file.ai/prod/v1, or your regional host. Personal (multi-project) keys cannot call this endpoint; use a project key from a project admin.

Understand the result

The update returns:

Response

{ "success": true }

Read the defaults back with GET /v1/project/setting:

Response

{
  "projectId": "5f5f726eea75272d54e1e1e1",
  "ocrModel": "Beethoven_ENG_GP25.2",
  "formTargetLanguage": "English",
  "isSplit": false,
  "createdAt": "2021-01-01T00:00:00.000Z",
  "updatedAt": "2021-01-01T00:00:00.000Z"
}

Field	App setting
`ocrModel`	AI Parsing Model
`formTargetLanguage`	File Target Language
`isSplit`	File Splitting

The other settings, including schema locking and classification, are app-only at project level.

Next steps

OCR models

Every AI parsing model and when to use it.

AI Schemas

Manage the schemas in the Extraction Schema tab.

Reference

Tabs

Tab	What it holds
General Settings	How documents are processed.
Extraction Schema	Every AI Schema in the project, and the Schema Locked toggle. See AI Schemas and Schema locking.
Notification	Email alerts about processed files. See Notifications.

General Settings

Setting	Options	Default
File Splitting (`isSplit`)	Do not split: the file is processed as one document. Smart splitting: a file that contains several documents is split into separate documents.	Do not split
Excel / CSV File Splitting (`isSplitExcel`)	Do not split: the workbook is processed as one document. Smart splitting: each worksheet becomes its own document.	Do not split
File Target Language (`formTargetLanguage`)	As-is (`as_is`), or a specific language.	As-is
AI Parsing Model (`ocrModel`)	The OCR model that reads document content, such as `Beethoven_ENG_GP25.2` for general documents or `Beethoven_Direct_Form_Filling (GP2.5)` for form workflows. See OCR models.	`Beethoven_ENG_GP25.2`, unless your project was configured otherwise
ML/AI-based Classification	Classifies documents automatically during processing with a pre-trained model.	Active

isSplitExcel can be set per upload but not with PATCH /v1/project/setting; change the project default in the app.

Get started

Products

Build

Go live

What it does

When to use it

How it works

Try it

Understand the result

Next steps

OCR models

AI Schemas

Reference

Tabs

General Settings

​What it does

​When to use it

​How it works

​Try it

​Understand the result

​Next steps

OCR models

AI Schemas

​Reference

​Tabs

​General Settings

What it does

When to use it

How it works

Try it

Understand the result

Next steps

Reference

Tabs

General Settings