The challenge
In credit origination, the payment capacity analysis depends on the documents the applicant provides: bank statements, pay stubs, employment certificates. Every file arrives in a different shape — PDFs from different banks, scanned pay stubs, photos. An analyst reviews them by hand, transcribes figures and consolidates income and expenses. Getting a file ready can take days, and missing documents are detected late — when returning the case has already stalled the application. Meanwhile, the applicant waits.
The proposed solution
A flow that structures the file before it reaches the analyst:
- AI extraction of bank statements and pay stubs, with no templates per bank or employer.
- Automatic consolidation of the applicant’s income and expenses into a single view.
- File validation before the analyst: completeness, validity dates and consistency across documents.
- The analyst receives the case with the supporting documents already structured and the alerts flagged: they decide, instead of transcribing.
Expected results
Preliminary figures, subject to the ongoing validation:
- Documentary payment capacity analysis: from days to minutes.
- 100% of files with supporting documents validated before reaching the analyst.
- Fewer returns due to missing or expired documents, and a faster credit decision for the applicant.
Status
This concept is under validation with real credit origination operations. The metrics are design targets and will be confirmed with measurements on real cases.