Understanding Expense OCR

Emburse Enterprise’s OCR (Optical Character Recognition) engine for Expense quickly and accurately captures data from uploaded images to automatically populate it in expense reports, eliminating the need for manual data entry.

For more details on Emburse's AI-powered OCR transcription platform, see New OCR Transcription Platform FAQ.

Data Extracted

Data is extracted from receipts via the Emburse Transcription Platform (ETP), which uses Machine Learning-based processing to identify the Expense Type and capture the following fields from uploaded receipts:

• Transaction Date*
• Amount*
• Currency Code*
• Merchant
• Last 4 Digits of the Credit Card
• VAT Information

* In the event that the values of these required fields are unable to be extracted by ETP, the system will automatically populate them as follows:

• Transaction Date: Date the receipt was uploaded
• Amount: 0.00
• Currency Code: The expense owner's primary currency

For information on how to map OCR category descriptions to expense line-item types in order to automatically assign the correct categories when a receipt is uploaded to Emburse Enterprise, see "OCR Receipt Expenses" in Assign Expense Types Admin. 

Optimized Languages

Although Emburse Enterprise's OCR for EXPENSE can capture data from most languages, it has been optimized for the following languages:

• English
• Spanish
• German
• Dutch
• French
• Italian
• Portuguese
• Japanese

Emburse Enterprise will continue to optimize the Machine Learning model for more languages in the future.

Configure via Customer Preferences

Administrators with appropriate permissions may configure OCR via the Customer Preferences admin screen. For complete details, see "Enable Expense OCR" and "Prevent Splitting of Images with Multiple Receipts" in the Expense section of Customer Preferences Admin Screen.