PERSON OF THE WEEK: AI-powered software solutions are seeing increased adoption across the mortgage industry, as lenders and servicers see the operational gains from intelligent automation.
Mortgage software firms are increasingly bundling AI into their core offerings, accelerating the proliferation of AI across the industry.
In a recent interview with MortgageOrb, Kayla Eames, product manager for Santa Clara, Calif.-based software firm Tavant, shared her views on how AI technology will help improve the mortgage underwriting process and why wider adoption is inevitable.
Q: How will AI technology change the mortgage landscape in the near future?
Eames: AI technology is rapidly transforming the mortgage landscape by automating and streamlining processes, accelerating decision-making, enhancing risk assessment by leveraging this technology to perform clerical and review tasks, all while personalizing the borrower experience. AI can save valuable time and resources by analyzing data, classifying and extracting documents, and providing real-time feedback on documents as well as analyzed data.
It is projected that more than half of lenders will deploy AI-powered solutions by the end of 2025. Anticipated advancements for 2026 include voice-activated mortgage applications, even more sophisticated predictive models for pricing and risk, and a continuous data feedback loop that improves every stage of the mortgage lifecycle.
AI is moving mortgage lending toward a more efficient, inclusive, and customer-centric model thereby achieving faster approvals, superior risk management, and a significantly improved experience for both the borrower and the lenders.
Q: AI Agents are a hot topic currently. What are AI agents, and how will they impact mortgage lending?
Eames: AI Agents are digital workers that use machine learning, business rules, and real-time data to take actions, understand context, use external tools, as well as to utilize reason and self-correcting behaviors without human intervention.
AI Agents will impact mortgage lending by reducing loan processes from hours or days to only minutes. AI Agents can help to enhance the borrower experience by answering questions, sending status updates, pre-qualifying borrowers, collecting documents, clarifying conditions or explaining complex information.
Additionally, AI agents can also provide support to borrowers and mortgage lending professionals 24 hours a day, 7 days a week.
Q: How will Agentic AI impact mortgage lending?
Eames: Agentic AI will help to reduce operational costs, decrease manual labor, accelerating mortgage cycle times, improving compliance and loan quality, enhancing the experience for the borrower, and reducing fraud risk. AI Agents can provide support to borrowers each step of the way through the loan application process, which will help minimize the risk of borrowers getting confused or frustrated, and help to ensure that they complete the loan application.
Additionally, Agentic AI will be able to analyze the loan data and documents immediately as it is collected from the borrower, thus expediting the approval process from days to minutes.
Q: How do you envision Agentic AI systems changing the day-to-day workflow and operations for an underwriting team – not just in terms of speed, but in how underwriters spend their time and the types of decisions they’re empowered to make?
Eames: Agentic AI systems will fundamentally reshape the underwriting team’s daily work not only by accelerating tasks, but also by reimagining what underwriters focus on and empowering them to make higher-value, more strategic decisions without the need for manual work as analysis will be automated for income, assets, credit, collateral, etc. Underwriters will gain efficiencies and then have time to focus efforts on complex loan situations that require knowledge, experience, empathy, and sound judgement.
Additionally, underwriters will have the opportunity to focus on becoming experts in other loan products that they may have not had the time previously (example: bond loans, non-QM, or jumbo) to master or get their respective designations (LAPP, SAR, or DE).
Q: What new risks does Agentic AI introduce to the underwriting process, and how can underwriters adapt accordingly?
Eames: Agentic AI introduces several new risks to the underwriting process and there are likely to be many regulatory questions from various parties like investors or state or federal regulators. Underwriters can prepare and adapt by embracing the technology with an open mind, integrating oversight, perform robust auditing procedures, and continuously evaluate their workflows such as doing side-by-side comparison tests.
For example, an underwriter can manually review the loan file, then review the automated conditions subsequently to ensure that AI yielded the same expected results. This will help the underwriter to gain confidence with the technology while they build muscle memory reviewing the AI feedback. Ultimately, it will equip the underwriters and leaders to have both a full understanding of the implications of Agentic AI in the underwriting process while preparing them with the answers that could be asked from a regulatory perspective.