Articles in this section

Which data are automatically extracted from the OCR?

Artificial Intelligence, making your life easier


Most of the the time, we use the Automatic mode when creating an expense in Jenji.
Data are automatically extracted from the receipts by our OCR recognition system, called ADDL. 

Thanks to our OCR - one of the best recognition accuracy and speed on the market - the main fields of your expenses are filled in automatically, such as the seller, the amounts or the category. 

What about electronic invoices in PDF format?


ADDL works perfectly with different types of receipt: restaurant bill, hotel invoice, purchase receipt, etc.  

Our OCR extracts also data from PDFs, no matter which platform is used, web or mobile (iOS and Android).

Data will be extracted entirely (Seller, Date, Total, Amount, Currency, Category) from official electronic invoices for the following providers:

  • Transport: Uber, G7, Chauffeur Privé,, Thalys, Trainline, Captain train, Eurostar, Easyjet
  • Telecom: Orange, Sosh, Bouygues Telecom, SFR, Free mobile, Free Internet
  • Accommodation: Airbnb
  • Other: LinkedIn, Apple, Amazon

Note: The payment method has to be entered manually in the case a payment method by default has not been configured yet. 

What are the results on foreign receipts?


According to the receipt, a manual input might be needed.  

Even though the OCR brings support and convenience, human intervention is required when science reaches its limits. 

The OCR is always right, false!


Data extraction can be in some cases incorrect or missing. 

Results depend on a multitude of factors that are difficult to control: phone model, quality of the picture, receipt format, etc. 

Being aware that no OCR can be 100% accurate, our focus is on extending the list of providers for which we support the format of their official electronic invoices (PDF). 

Was this article helpful?
0 out of 0 found this helpful