BNP Paribas Unveils AI-Powered Solution to Expedite Mortgage Approvals
BNP Paribas is leveraging artificial intelligence (AI) to streamline its mortgage process, aiming to reduce processing times and improve customer satisfaction. According to Sébastien Perrigault, Head of Retail Financing at Commercial & Personal Banking in France, the new system has already shown promising results.
“Obtaining a mortgage is one of the first steps [in a property purchase],” Perrigault explained. He noted that it can often be a source of stress and delays for customers. To address this, the bank has implemented an intelligent document processing (IDP) solution, which began in November 2024.
The goal of the AI integration is to optimize the mortgage process by automating several key steps. The IDP, which utilizes NLP (Natural Language Processing) technologies, is designed to read customer-provided documents, extract relevant information, and verify its accuracy. This helps reduce the workload on advisors and minimizes errors.
“More than 80% of the document data is automatically read by AI and 60% of data is automatically matched and validated by AI,” Perrigault said. This automation leads to an estimated time savings of around 30 minutes per file.
Key Features and Benefits:
- Faster Processing: The IDP solution has reduced the average time to issue a digital mortgage offer.
- Reduced Errors: AI verifies the consistency of information, ensuring accuracy between various documents.
- Time Savings: Advisors now save approximately 30 minutes per file. Overall, the IDP is intended to reduce the time for mortgage approvals by five days, with the potential to grant a mortgage in less than 30 days.
- Enhanced Customer Experience: The technology contributes to a more straightforward and expeditious journey for customers.
Technical Details
The AI system leverages a combination of classic AI and the bank’s internally developed IDP software, called Doc Factory. The IDP software is based on NLP technologies and has been trained on thousands of documents to identify and verify data.
The initial focus is on core mortgage applications, including the four prime documents: the purchase agreement, property tax, pay slip, and tax notice. BNP Paribas plans to expand the system to include additional documents in 2025, such as energy performance certificates and savings account statements. The bank aims to issue approximately 20% of offers electronically, without manual intervention, beginning in 2026, further reducing processing times and costs.
“For our advisors, IDP acts as a ‘super assistant’ at several stages of file processing,” Perrigault stated. He highlighted the close collaboration between different teams as a key factor in the project’s success.
By embracing AI, BNP Paribas is working to make the mortgage process faster, more accurate, and more satisfying for both its advisors and its customers.
