Underwriting Using AI and ML with Legacy Loan Origination Systems (LOS)





Lending has always been a major business in North America and, over the years, every disruptive technology has contributed to enhancing this sector. Since there are various steps related to funding mortgages, underwriting is probably the most elementary and complex, consuming plenty of man-hours. With the power of Artificial Intelligence (AI) and machine learning (ML), the entire underwriting process can be automated.

It is understandable that some big players in the mortgage and lending industries have legacy loan origination systems (LOS) and thus, they cannot just replace them all at once. However, if we start providing solutions to augment the outdated platforms as ancillary services to support the system, we can begin automating processes without the frustration and effort of replacing their current LOS.


Why Is It Important to Bring Intelligence in the Underwriting Process?

Lending is a data-intensive process and we can all agree that humans are much less efficient than algorithms and computers when it comes to quick data-processing and analysis.

Moreover, while deciding whether a loan application should be accepted or not, the creditworthiness of applicants is a major concern that must be determined by evaluating their documents and figures. Once you have all the information for an individual, you can simply tally the data you have for previous customers and evaluate the debt servicing behaviour. On the basis of this analysis, the actual creditworthiness is determined.

It is also worth mentioning that apart from just analyzing the data submitted by the applicant, a lender must verify their identity as well, due to regulatory standards. This is yet another area addressed by automated underwriting tools and makes the process much more efficient.

Therefore, if machine learning is deployed to automate this specific area, it can process more applications in the same amount of time and also ensure authenticity of analysis by removing the probability of errors.

Since the lending industry contributes majorly to the overall economy of a country and several business sectors are associated with it, the use of AI can further streamline the internal and external process of an organization.

Apart from this, an automated underwriting process also provides better feasibility for tracking the documents when needed, instead of just searching for the unorganized files in your system. Everything is arranged by default.


Since most of the legacy loan origination systems are quite large, complex and technologically outdated, it is not possible to replace them all at once as it might affect their interoperability. Therefore, one of the best approaches, in this case, is to take a part of the overall system and introduce AI into it to decrease the processing backlog. Since underwriting consumes the most time in the lending process and requires plenty of analysis, it is a good idea to use an AI-based underwriting tool (such as the one provided by FundMore) and plug it into your LOS.

Tags: mortgages, Automation, lending, Mortgage Processings