With Optical Character Recognition (OCR) we could already retrieve data from standardized documents such as a passport or an employer’s statement via pattern recognition and then save this data separately. This saved a lot of manual work, for example in the mortgage process. The shortcoming of OCR, however, is that it only works with a fixed layout. If the image is shifted a few pixels, the software no longer works. In addition, OCR is expensive and time-consuming.
A more efficient mortgage process
VisionAI solves these objections. It extracts data from documents with all kinds of formatting and can also perform checks at the same time. For example, Vision AI can verify whether a passport is still valid or check whether the salary on a bank statement matches the amount on the payslip. The mortgage process in particular, where a consumer has to submit many different documents, can be made much more customer-friendly with VisionAI. If you can perform automatic checks early in the mortgage process, this shortens the processing time and reduces dropouts. This prevents situations in which a consumer has had to wait so long that, for example, the employer’s declaration has already expired – with all the frustration that entails. And by automating the acceptance rules of mortgage providers, you can use VisionAI to predict early in the mortgage application process whether a consumer will receive a mortgage. Acceptance rules sometimes change, which quickly creates information asymmetry in the process. With business ruling, you can overcome this and keep the system up to date more easily.
Safely to the market
I am proud that we have been able to bring a technology like VisionAI, which has only been around for a year, to the mortgage market so safely. We have succeeded because we know the financial world well, because we understand customers and their processes well and because we recognize opportunities and can actually realize them. With the introduction of this technology, we are making a positive impact on many operational processes. This is currently the case in the mortgage sector, but this is only the beginning. VisionAI is particularly advantageous when there are high volumes of documents, when checking these documents requires a lot of manual work and when there is a great variety in how the documents can look (such as contracts, annual accounts and pay slips). And these situations occur in many places in the financial sector.
Simple compliance in asset management
I see promising applications of VisionAI in the asset management sector, for example around compliance. Every supplier has different conditions when it comes to data use. That makes it difficult to demonstrate that you, as an asset manager, meet those conditions. With VisionAI, you can extract data from data contracts and put it in a compliance system. Or think of the know-your-customer processes. With VisionAI, you can easily demonstrate that the required documents have been checked.
Value transfer in pensions
Or look at the value transfer in pensions. When merging pensions, the new administrator receives quotes from the deferred funds. The variables in those quotes differ from each other. With VisionAI, you can read those documents and merge the data. This is a solution, especially for smaller funds and insurers for whom it is too expensive to build expensive integrations themselves.
Transition communication regarding pension fund entry
When entering, pension funds send letters to participants with a range of numbers – considerably more than with the UPOs. It is of great importance that all these numbers are correct for each participant in each letter. With VisionAI, pension funds can ensure at the back end of the process, i.e. before sending out the letter, that all these numbers are actually correct and in the right place in the letter. A 100% final check.
Complete screening of portfolios
It also makes sense for asset managers who take over mortgage portfolios from each other to screen their new portfolio with VisionAI. This is often still done manually via random samples. That is time-consuming, expensive and not watertight. With VisionAI you do check the entire portfolio before the takeover and as an asset manager you know better what you are buying.
Technology and the human eye
VisionAI shows its value especially at high volumes because the technology takes over the repetitive control work and ensures more efficiency in processes and a higher quality of controls. Then it is up to the expert (the mortgage advisor, the acceptor, the asset manager, the pension specialist) to come up with advice or make the decision. But the impact of VisionAI on high-volume processes in the financial sector will be undeniably large.
