In 2024 alone, 1 in 134 mortgage applications contained fraud.
As mortgage fraud detection in primary and secondary markets relies heavily on human judgment, it can be difficult to detect, especially considering the volume of applications each underwriter must oversee on a daily basis.
However, Tikhnadhi Kamlakshya, a senior manager at Citizens Bank, believes that AI technology can be used to rectify this issue. AI tools, such as multimodal language models, can quickly identify irregularities and points of concern within applications, and integrated telephony and data systems can enhance tracking, transcription, and analysis in mortgage operations, making it much easier to identify fraud.
This article will explore the various ways in which AI and cloud-based tools can be used to revolutionize mortgage operations.
The Problem
Challenges Faced by Underwriters
The most significant problem with the current methodology for identifying mortgage fraud is that it relies on human judgment. When underwriters are tasked with manually overseeing numerous applications in a single day, they simply do not have the time or the bandwidth to check each application as carefully as necessary. This means that they often fall victim to oversight, with fraudulent applications slipping through the cracks.
For example, they may fail to identify whether or not the data in the application is inflated, which impacts a lender's loan eligibility and leads to complications in the future.
Not only that, but relying on manual processes slows down an institution's operations considerably, leading to frustration among both employees and potential lenders. This could also result in them struggling to prioritize high-value loans efficiently due to the volume of work they have in front of them, inadvertently putting their company's finances at risk.
Operational Impact
Another common challenge underwriters face is being overburdened or overworked (i.e., processing applications manually), which can lead to significant delays.
For example, while the majority of loan applications are processed within two to six weeks, this is not always the case, reducing both profitability and customer satisfaction rates and leading to a range of challenges for mortgage lenders and institutions.
Introducing Tikhnadhi Kamlakshya
Tikhnadhi Kamlakshya, a Sr. Manager of Cloud Mix – Innovation & Modernization at Citizens Bank, has over 20 years of experience in technological adoption across numerous industries, including the financial services sector. He hopes to solve many of the issues associated with mortgage operations using the latest AI technologies.
Kamlakshya led the transition from an outdated legacy system to Salesforce and Marketing Cloud within the mortgage sector (retail BTC) of Citizens Bank using AI. This shift quadrupled underwriter productivity, allowing them to complete 60–80 applications per hour instead of just 15. They were able to process over $1 billion in annual loan value, showcasing the transformative power of cloud and AI technology.
This project, alongside the rest of the work carried out during Kamlakshya's four-year tenure at Citizens Bank, has contributed to the bank being consistently ranked in the top 10 of the national mortgage sector, as recognized by National Mortgage News.
Another of his successful projects came while working as a Salesforce Consultant in the State of Michigan, where he led the implementation of the Salesforce system, making them the first agency in the state to adopt a cloud-based integration.
Leveraging AI-Powered Fraud Detection
Tikhnadhi hopes to use AI as a tool to identify fraudulent activity, such as inflated or inaccurate depictions of income and mortgage eligibility.
By using AI-powered tools such as machine learning models and multimodal data inputs to identify fraud, the pressure faced by underwriters decreases significantly.
This means that they can focus on high-value tasks, such as customer relationship management and regulatory compliance, as they're no longer spending so much time conducting routine checks. They can also rank applications based on risk.
A Future-Ready Banking Ecosystem
Tikhnadhi predicts that AI technology will redefine key operations within the banking industry, fueling growth and innovation across adjacent financial sectors, including mortgage lending.
For example, by introducing and developing AI-driven systems specifically for underwriters, Tikhandhi hopes to overcome many of the challenges faced by even the most experienced underwriters, especially regarding identifying fraudulent activity. When implemented seamlessly, these tools can quickly pick up on data irregularities or inconsistencies within applications so that underwriters can act accordingly.
This prevents significant institutional losses while also ensuring that they can process a larger number of applications thanks to reduced processing times.
Furthermore, Tikhnadhi believes that AI can be used to revolutionize how customers engage with their bank or banking system. Right now, he's working on a project that is set to reinvent the banking referral system at Citizens Bank. Once introduced, these tools can be used to analyze user spending habits and credit profiles, which means that banks will be able to provide clients with specifically tailored advice and guidance moving forward.
By taking a more personal approach to the services they offer, banks and financial institutions can strengthen their relationship with their clients while also improving their standing within the industry. {Add Researchgate and AI Business use-cases Links}
There are many scenarios in which AI technology can be effectively utilized in banking, especially when it comes to mortgage applications. After all, it's a relatively easy way to deal with many of the pain points that underwriters are currently facing, such as being overwhelmed and having difficulties identifying fraudulent activities before it's too late.
To learn more about Tikhnadhi Kamlakshya and his work, read his published articles focused on AI business use cases for various industries.
