Custom Document Type/Fields

PreviousNext

Learn how to create custom document types and add custom fields

Custom Document Types Overview

A custom document type is a document schema that you create to define the structure of specific types of financial documents. While 1Flow comes with built-in document types (like Invoice, Receipt, Bank Statement), you can create custom document types for documents that aren't covered by the defaults.

Custom document types allow you to:

  • Define exactly which fields should be extracted
  • Organize fields into header and line item collections
  • Configure how AI extracts each field
  • Create document types specific to your industry or workflow

Custom Fields

Custom fields are data fields that you add or modify in document types. Importantly, all document types and fields are customizable - there's no difference between system (built-in) document types and custom document types in terms of customization capabilities.

System document types (like Invoice, Receipt, Bank Statement) are simply pre-created for convenience, but they're meant to be customized to your organization's specific needs. You can:

  • Add new fields to any document type (system or custom)
  • Modify existing fields - change descriptions, extraction hints, flags, and other properties
  • Remove fields that aren't relevant to your workflow
  • Customize field behavior - adjust AI extraction, formulas, display options, and more

Every field, whether it came with a system document type or you created it yourself, can be fully customized per organization. System document types are just starting points - they're not locked or special in any way.

Custom fields screen
Custom fields screen

Creating a Custom Document Type

Step 1: Create the Document Type

  1. Navigate to Document Types in the main navigation
  2. Click Create Custom Document Type
  3. Fill in the basic information:
    • Name: Technical identifier (e.g., UTILITY_BILL)
    • Display Name: Human-readable name (e.g., Utility Bill)
    • Description: Optional description for AI classification
    • Color: Visual color identifier

Step 2: Define Header Schema

Header fields are extracted from the document header and apply to the entire document:

  1. Click Add Field in the Header section
  2. Configure the field properties (see Field Configuration below)
  3. Repeat for all header fields you need

Common header fields:

  • Vendor/Provider name
  • Document number
  • Date
  • Total amount
  • Currency
  • Due date
  • Account number

Step 3: Define Line Item Schemas

Line item fields apply to individual items within collections:

  1. Click Add Collection to create a new line item collection
  2. Give the collection a name (e.g., items, charges, transactions)
  3. Click Add Field within the collection
  4. Configure each field

Common line item fields:

  • Description
  • Quantity
  • Unit price
  • Amount
  • Category
  • SKU/Product code

Step 4: Configure Extraction

For each field, configure how AI should extract it:

  • Description: Explain what the field represents
  • Extraction Hint: Provide specific guidance for AI extraction
  • AI Extractable: Enable/disable AI extraction for this field

See AI Custom Fields for detailed information.

Step 5: Test Your Schema

  1. Upload a sample document
  2. Review the extraction results
  3. Adjust field configurations as needed
  4. Iterate until extraction quality meets your requirements

Field Configuration

Each field has several configuration options:

Core Properties

  • Field Name: Display name shown in the UI
  • Technical Name: Internal identifier (used in API and formulas)
  • Field Type: Data type (text, number, currency, date, boolean, enum)
  • Enum Values: For enum fields, define allowed values

Flags

  • Hidden: Hide field from UI (useful for computed fields)
  • Required: Field must be present (for validation)
  • AI Extractable: Allow AI to extract this field (disable for workflow fields, notes, or formula fields)
  • Searchable: Include in search index
  • Is Important: Mark as important field
  • Is Description: Use as primary description field
  • Is Reconcilable: Allow reconciliation on this field
  • Is Institution: Mark as institution identifier
  • Is Primary Key: Use for dataset upsert operations

See Non-AI Fields for information about when to disable AI extraction.

Search and Display

  • Pinned: Pin column to left or right in tables
  • Currency Field: For currency fields, reference the currency code field

Formulas

  • Sum Fields: Auto-aggregate line item values
  • Code Fields: Custom JavaScript for computed values

See Formula Fields for detailed information.

Field Types

Text

Free-form text fields for names, descriptions, notes, etc.

Use cases:

  • Vendor names
  • Product descriptions
  • Notes and comments
  • Reference numbers

Number

Numeric values without currency formatting.

Use cases:

  • Quantities
  • Percentages
  • Counts
  • Measurements

Currency

Monetary amounts with currency support.

Use cases:

  • Amounts
  • Prices
  • Totals
  • Fees

Configuration:

  • Link to a currency field to specify the currency code

Date

Date values in YYYY-MM-DD format.

Use cases:

  • Invoice dates
  • Due dates
  • Transaction dates
  • Expiration dates

Boolean

True/false values.

Use cases:

  • Flags
  • Status indicators
  • Checkboxes

Enum

Predefined list of allowed values.

Use cases:

  • Status (PAID, UNPAID, OVERDUE)
  • Categories
  • Types
  • Priorities

Configuration:

  • Define allowed values when creating the field

Best Practices

Naming Conventions

  • Technical names: Use camelCase (e.g., invoiceNumber, totalAmount)
  • Display names: Use proper capitalization (e.g., Invoice Number, Total Amount)
  • Keep names descriptive: Avoid abbreviations unless widely understood

Field Organization

  • Group related fields: Organize fields logically in the schema
  • Use consistent naming: Follow patterns across your document types
  • Document field purposes: Use descriptions to explain field usage

AI Extraction

  • Provide clear descriptions: Help AI understand what to extract
  • Use extraction hints: Give specific examples or format requirements
  • Test thoroughly: Upload sample documents to verify extraction quality

Performance

  • Limit searchable fields: Only mark fields as searchable if needed
  • Use formulas for computed values: Don't extract what can be calculated
  • Hide unnecessary fields: Use hidden flag for internal-only fields

Examples

Example: Utility Bill Document Type

Header Fields:

  • providerName (text) - Utility provider name
  • accountNumber (text) - Customer account number
  • billDate (date) - Bill date
  • dueDate (date) - Payment due date
  • totalAmount (currency) - Total bill amount
  • currency (text) - Currency code

Line Item Collection: "charges"

  • description (text) - Charge description
  • amount (currency) - Charge amount
  • category (enum) - Charge category (ELECTRICITY, WATER, GAS, etc.)

Example: Purchase Order Document Type

Header Fields:

  • vendorName (text) - Vendor name
  • poNumber (text) - Purchase order number
  • orderDate (date) - Order date
  • totalAmount (currency) - Total order amount

Line Item Collection: "items"

  • sku (text) - Product SKU
  • description (text) - Product description
  • quantity (number) - Ordered quantity
  • unitPrice (currency) - Unit price
  • amount (currency) - Line total (formula: quantity × unitPrice)

Next Steps