By Asheesh Mehra, Group CEO and Co-founder, AntWorks

As seen in many other countries, the COVID-19 pandemic sent a shockwave through the UK housing market, bringing the entire sector to a virtual standstill. As lockdown restrictions  ease, we are now witnessing a housing boom like no other, as thousands have entered the market looking to capitalise on the UK government’s new stamp duty relief on properties priced up to £500,000. At the same time, however, the economic fallout from this financial crisis has resulted in almost 750,000 people losing their jobs and countless more being furloughed, leading to an increase in property remortgaging requests and payment holidays.

As a result, banks and mortgage companies now find themselves inundated with new mortgage applications, refinancing and forbearance applications. To support this, there is now a drastic need for increased manpower to manually process the plethora of mortgage enquiries in a more efficient manner. That being said, the uncertainty of future pandemic impact and restrictions being imposed at a local or global level is leaving the industry under severe pressure to deal with the demand as quickly and effectively as possible.

Like many other industries feeling the impact of the COVID-19 crisis, the mortgage sector needs to deploy digitisation in their operations in order to scale their business faster than before or risk being left behind. Artificial Intelligence, deployed in conjunction with intelligent automation, can help ease the burden on mortgage brokers and lenders by accelerating the mortgage loan process and reducing costly errors caused by manual input.

Achieving speed and scale through intelligent automation

Automation is a viable and logical  solution for mortgage lenders as approximately 60 – 70 per cent of tasks in mortgage processes across the value chain are, replicable and prep-based tasks that are suitable candidates for automation. Traditionally, mortgage companies frequently conduct borrowing assessments that require careful analysis and comparison of customer data. This includes checking and establishing customer credit history as well as affordability by manually processing data from income documentation such as bank statements and payslips. This is a tedious but highly necessary process known (rather un-fondly) in the industry as the “stare and compare” stage of mortgage processing

These tasks require a huge amount of paperwork and form filling, which is not only time-consuming but also prone to human error. Furthermore, in their day-to-day routine, mortgage processors are required to literally unpackage and organise documents that are often in paper formats. This is a laborious process especially when executed across multiple mortgage applications at the same time.

This is where intelligent automation steps in to help mortgage companies take on and complete far more work, at a much faster rate and with higher accuracy. Automation can relieve mortgage workers from repetitive tasks such as manually populating the loan origination systems. This means that customers can get loans approved quickly and efficiently. In fact, a global mortgage provider leveraged the power of automation to increase the speed at which mortgage documents were being generated by up to 90 per cent without compromising the integrity of its review process. What’s more, they also managed to improve the turnaround time for the more complex documents by 100 per cent.

Cognitive Machine Reading (CMR) based solutions are the answer for companies looking to achieve straight-through processing for all their mortgage documents. CMR enables mortgage companies to overcome the challenges of digitising unstructured data and achieve faster ROI with higher accuracy with data certainty. Additionally, it can help mortgage companies to cut down on loan processing costs by up to two-thirds and mortgage origination time by 50 per cent.

The (fractal) science behind CMR is that it uses integrated AI capabilities to process highly complex unstructured data along with the more basic data formats. This data can then easily flow through the entire organisation via an end-to-end process achieved with little to no human interference.

Inevitably, all business data needs to be digitised so that it can feed analytics, drive automation, and provide those much-needed customer insights. CMR can play a part to eliminate repetitive and error-prone stare-and-compare tasks that are often a commonplace in mortgage processing. It is able to identify and process the context of data and validate it against existing information elsewhere. As a result, this speeds up the overall processing time for new mortgage and refinancing requests.

Avoiding common automation mistakes

Before kickstarting any digital transformation journey or automation projects, it is imperative that businesses look into avoiding the pitfalls of adopting the wrong automation tools. For example, utilising traditional Optical Character Recognition (OCR) technology for business processes can lead to significant data challenges which will slow down and impede automation goals. OCR is a simple data ingestion tool that is limited to only processing and automating structured data that comes in the form of fixed-field text. Given that 80 per cent of the data within most organisations is unstructured or does not have a predefined format (e.g. emails, images, signatures, social media feeds), OCR technology cannot ingest the vast majority of data trapped inside a mortgage process (or any other business process). In order to overcome this and improve its business process outcomes, one leading Insurance provider managed to process large volumes of unstructured data via CMR automation to achieve 75 per cent reduction in the manual data extraction of handwritten documents. Additionally, the company also achieved more than 92 per cent accuracy in identifying and processing handwritten content.

Critical, everyday business data contained in multiple formats such as emails, images, and handwritten material make up a large part of unstructured data. This is why businesses need to put greater emphasis on researching and identifying intelligent automation solutions that can unlock this date to achieve their business goals. CMR enables mortgage companies to significantly accelerate the course of identifying and classifying all types of documents by cutting down the reduction time for processing mortgage claims by 90 per cent with a substantial level of accuracy (75%). What’s more, it enables any organisation to automate at scale, bringing true automation as a company-wide approach rather than a segregated one.

The COVID-19 pandemic has managed to speed up the need for businesses to embrace digital transformation. This may well be the catalyst for many mortgage organisations steeped in antiquated legacy-based ways of working to refine and streamline their business operations via straight-through processing. It is clear that companies can successfully automate entire business operations to not only improve their operational efficiency but also achieve organisational resilience in a long run. And the faster mortgage lenders can tackle their processes right now, the better for the sooner they can pass those efficiencies and savings onto customers to help rebuild the economy and bolster the housing market in the UK and elsewhere.