PERSON OF THE WEEK: For the mortgage lending industry, artificial intelligence (AI) holds great promise to reaching the end goal of a fully digital, fully automated, fully compliant home buying process that takes only minutes and yet is personalized and seamless for the borrower.
That day is coming soon, says Sandeep Shivam, AMP, product evangelist and associate director with Tavant, in a recent interview with MortgageOrb.
Q; How is technology transforming the mortgage industry as a whole?
Shivam: Over the last 30 to 40 years, the mortgage industry has experienced significant technological advancements. We’ve progressed from basic digital record keeping to fully digitalized application processes, and today, cloud-based infrastructure is helping reduce IT costs while improving scalability.
These innovations have undoubtedly increased convenience for borrowers. Automated verification services have streamlined document collection and validation. Operational teams have benefited from tools like OCR technology to process information more efficiently. However, despite these advances, the overall loan cycle time has remained largely unchanged since the 1990s. In fact, the cost of originating a mortgage has continued to rise over the years, even with these efficiencies in place.
We are now at a critical inflection point, standing on the threshold of what could be the industry’s next major transformation—Artificial Intelligence (AI). In conversations with industry leaders, there is both excitement and optimism around AI’s potential. The mortgage process is highly human-centric, involving underwriters, loan officers, processors, and other stakeholders. AI’s ability to replicate certain decision-making processes and automate complex workflows presents an opportunity to significantly reduce the human effort required.
Q: How are borrower expectations shifting in response to an increasingly digital and self-serv mortgage landscape?
Shivam: In the past, borrowers primarily sought the best rates and price, paired with excellent service. Later, their expectations expanded to include the convenience of a digital application process. Today, those expectations have evolved far beyond that.
Modern borrowers demand seamless multi-device support, allowing them to start an application on a laptop and continue on a mobile device – whether through an app or a browser – without losing any progress. They expect instant updates on every action they take, along with the ability to access help instantly at the click of a button. Transparency is non-negotiable; borrowers want real-time visibility into the status of their application.
They also value a personalized experience – one where they feel treated as a top priority, with interactions tailored to their specific needs.
Most importantly, security and data protection have become critical expectations. In a world where cyberattacks are increasingly common, borrowers expect lenders to safeguard their sensitive personal and financial information with the highest standards of security.
Q: What distinct advantages does agentic AI offer over traditional chatbots in enhancing borrower engagement?
Shivam: First, let’s understand the basic difference between Chatbots and agentic AI. Chatbots are automated solutions that operate primarily on predefined scripts and rigid rule sets. In contrast, an AI agent can take proactive action toward achieving a goal. It possesses reasoning capabilities – almost like having its own “brain” – allowing it to plan, adapt dynamically to the context, and select the most efficient path forward. When multiple AI agents work collaboratively, they create what is known as an agentic experience – a coordinated, autonomous system. Given these differences, incorporating agentic AI into a workflow offers several advantages:
- Hyper-Personalized Experience: Agentic AI can interact with the borrower or other users in a natural, conversational way, adjusting questions based on prior responses. It can intelligently call APIs to pull credit and property data, assess eligibility in real time, skip unnecessary steps, and present multiple relevant product options – all without human intervention. This creates a seamless, highly tailored borrower journey that feels both personal and efficient.
- Instant Decision-Making: Agentic AI has the capability to process information in real time, applying complex decision rules the moment data is received. This is especially valuable in the underwriting process, where it can evaluate submitted documents and data against both lender-specific policies and GSE guidelines. Leveraging this analysis, the system can automatically raise underwriting conditions when necessary and seamlessly clear them as soon as the required criteria are met.
- Increased Efficiency: Multiple AI agents can work concurrently, each specializing in a distinct domain—such as credit evaluation, collateral assessment, title verification, income and asset analysis, and pricing optimization. One agent may calculate debt-to-income ratios, another may assess property conditions, while a third determines the need for a letter of explanation based on credit and income reports. Coordinated division of labor speeds up the process, maintains consistent reviews, and results in efficient, accurate workflows.
Q: What steps must the mortgage industry take to enable real-time pre-approvals and truly instant mortgage closings?
Shivam: The mortgage industry has made strong progress toward real-time approvals, but we are still some distance away from achieving truly instant mortgage closings. Today, technology has enabled certain products – such as home equity loans and refinance transactions – to offer instant pre-approvals and disclosures. Solutions like Fannie Mae’s EarlyCheck within Desktop Underwriter (DU), combined with soft credit pulls, have made these capabilities practical. Even in purchase transactions, instant pre-approvals are now achievable, though instant disclosures are still a few steps away.
When it comes to instant mortgage closings, several fundamental changes are required. The most significant barrier lies in regulatory frameworks. At present, a licensed professional must still speak directly with the borrower. A human underwriter must review and approve the loan file, and property appraisals still require physical inspections. Additionally, the “seven-day rule” prevents closing a loan until at least seven days after initial disclosures have been sent. For instant closings to become feasible, these regulatory requirements will need to evolve – likely in parallel with technology proving itself to regulators as both reliable and compliant.
We could see a future where home buying is as seamless as purchasing a car online – where a borrower can secure financing for their dream home while sitting on their couch, discussing options with their family, and making informed decisions within minutes. Ongoing advancements in technology, coupled with greater industry trust, will make that vision a reality.