FINspeak

There have been numerous technology trends over the past several years; some turn out to be mere hype, while others deliver tangible returns on investment. Identifying the winners is a priority for all businesses as they evaluate where to focus spend.

This is especially true for financial sector businesses such as mortgage lending. Artificial intelligence (AI), natural language processing (NLP) and machine learning (ML) are some of the more recent technologies that are solving critical business challenges and delivering ROI to early adopters. But can the interplay of these technologies unlock even greater value?

Cloud + AI = Cognitive cloud lending

Marrying AI with the benefits of the cloud will create cognitive cloud platforms that can have far-reaching impact for mortgage companies. Cognitive cloud lending platforms and solutions can provide lenders, borrowers, brokers, dealers, real estate agents with more self-service and transparency options throughout the lending experience. Further, they can extend this ecosystem, making it possible to create and bring cognitive computing applications to the masses.

An apt example here would be the Mortgage Cadence Platform. It is the only comprehensive single system of record technology which is extensible through an open API layer, allowing banks/lenders to connect with third-party providers of their choice. It also offers valuation, verification, title, credit, fee, and compliance services. Another example would be Cloudvirga, the firm behind the cloud-based intelligent Mortgage Platform® (iMP). Designed to streamline the mortgage process, this platform is digitalizing the mortgage industry by deploying automated workflows to reduce cost, increase transparency and shorten the time it takes to close a loan for both borrowers and lenders.

Four ways to enable cognitive cloud lending

1. Replace monolithic software solutions - To get projects up and running, most large companies use monolithic IT architecture as it is faster to set up. However, as the system grows, the code and architecture become more complex and more developers are needed to maintain these. Companies also lose critical speed, flexibility and agility to respond to the market. Organizations embarking on digital transformation should consider an agile style of application architecture that enables rapid delivery of new cloud-based digital services.

Microservices architecture is one way of achieving this. This architecture encapsulates business entities that orchestrate multiple business entities as in the case of credit evaluation. For example, microservices can support and automate an existing loan origination system. It can accelerate risk decisioning processes in a secure manner whereby risk scores are calculated within a second of the application being received. The result? Applications can be approved in a fraction of the average 'time to cash'. To put this in context, the average time to cash is 40 days and even top performers take 18 days.

2. Encapsulate the core with process APIs - Instead of designing infrastructure around applications, service-oriented architecture (SOA) focuses on specific services. SOA is very useful in supporting business processes without having to worry about the underlying applications. Process APIs act as the communication layer, combining data from disparate sources and making it easily discoverable. For mortgage lenders, using process APIs brings about visibility, programmable services and greater agility.

Faced with numerous requests from customers looking to switch mortgages, a US-based home mortgage lender decided to automate its loan underwriting system. By leveraging SOA, the mortgage company can employ reporting tools and follow industry best practices, all while managing loan applications seamlessly.

3. Leverage cognitive BPM for intelligent automated services - Cognitive business process management (BPM) refers to the new BPM paradigm enabled by cognitive computing. It encompasses all the typical BPM aspects in addition to three transformational aspects: Firstly, CBPM will drive knowledge acquisition at scale, enabling knowledge-intensive processes (KiPs). Secondly, a 'plan-act-learn' cycle will emerge as the new process meta-model. Thirdly, the models can study process descriptions both during design and run-time. This will open up opportunities for new levels of automation and business process support for both traditional business processes and KiPs. A research paper on CBPM hypothesizes that cognitive computing will leverage AI to manage dynamic processes, achieving intelligent automation, continuous improvement and higher performance.

In the mortgage industry, cognitive BPM can be useful to determine customer satisfaction. Let's take an example, when a customer's loan gets approved, he/she is directed to the bank's loan servicing department. This department monitors changes to the payment plan and ensures proper payment collection. Its operations include outbound and inbound calls that generate call transcripts. Now, if the bank/lender applies cognitive analysis to this process, it can determine whether its employees are asking the right questions, how efficiently are they working, how polite are they with the customers, so on and so forth. Based on these insights, the bank can take actions to either improve customer communication or leverage new ways to improve customer satisfaction as well as the bank's profitability.

4. Connect front and back-end systems - To improve lending and advisory capabilities in future, banks will need to harness low-code applications and automation technologies that foster connections between front-end and back-end systems. This will also support smart case management, a capability that allows adaptive business processes In a post COVID-19 scenario, this kind of agility can mean faster loan disbursals for SMEs seeking financial stimulus from their respective governments. In the UK, nearly 69,000 loans were approved within 24 hours of a 100% government-backed scheme, highlighting the mounting pressure on loan processing teams. Using APIs to facilitate back-end integration promotes reusability as it unlocks new opportunities to integrate with other front-end technologies like chatbots.

Intellect Design Arena, a financial technology provider, built the Intellect Digital Lending (IDL) 20 on an 'always on, always aware' concept. Designed on a do-it-yourself (DIY) principle, IDL 20 allows banks to create their products anywhere, anytime. Banks can also make real-time informed credit decisions. It also enables banks to deliver a real-time solution to their customers by giving them access to a 360-degree view of their credit portfolio. Its fully automated robust architecture reduces operational costs by increasing efficiency.

Conclusion

The mortgage lending business needs to leverage new trends driven by technologies like cognitive cloud that can accelerate lending processes, making them intuitive, smart and relevant to customers. It is up to banks to align these technologies with the business and develop innovative lending models that leverage the power of intelligence.