Nowadays, car dealers invest a lot of resources in checking the Personal Income Tax (IRPF) data of potential customers manually. Apart from being expensive in terms of time and staff costs, this way of working may be inaccurate and even lead to fraud in the process of transcribing data.
Foqum’s data extraction tool for Spanish Inland Revenue Service (AEAT) tax forms provides fluency and accuracy while automating the entire process.
A leading global car sales company requires the extraction of the IRPF Form 100 data in a faster, more accurate and more secure way for the financing of its vehicles by dealers.
The data extraction tool must combine scans of Form 100 documents from different years, as dealers work predominantly with documents from the same year and the previous year. Firstly, detection of the IRPF tax year is required, as in different years some fields may vary, or there may be minor modifications in the format. In turn, those documents feature different structures with regard to the personal and financial data of the buyer, which may be compiled in plain text or in tables. Therefore, the OCR needs to be adapted to guarantee legibility.
The tool must support simultaneous reading of several Form 100 documents, as the financing of a vehicle may be performed by several people.
The tool developed by Foqum enables the company’s teams to instantly extract IRPF data.
The process is: (i) to enter the Secure Verification Code (CSV) in the tool to access the IRPF document from the Spanish Inland Revenue Service website; (ii) to use a verification field such as the applicant's National Identity Card (DNI) number to detect possible fraud; (iii) to export the information according to the business logic required by the company in a web application aimed at facilitating finance decision-making by dealers. The web tool enables the registration of individual users to launch those requests and view their statistics.
The tool has enabled the analysis of thousands of finance transactions with scanning errors in less than 0.2% of cases.