Mortgage servicers have spent the past few years operating in a forgiving environment. Rising home values, low default rates, and a steady economic backdrop have kept serious delinquencies in check. However, several converging pressures are raising a practical question inside servicing operations: are current systems prepared for higher volume? These pressures include elevated interest rates, rising insurance and property tax costs, recent FHA policy changes, and stagnant wage growth.
The question has new urgency. Recent FHA policy changes have narrowed options for struggling borrowers, including a 24-month restriction on repeat modifications. Combined with higher rates, rising HOA fees, and climbing insurance costs, the conditions for an uptick in defaults are forming. Most servicers still rely on manual, fragmented workflows that buckle under pressure.
Sapan Bafna, Founder and CEO of Outamation, a workflow automation company focused on the default side of servicing, has been working on this problem for most of his career. With nearly two decades in mortgage servicing, including a long tenure at CoreLogic (now Cotality), he has lived through the 2008 financial crisis, the national mortgage servicing settlement, COVID-related forbearance waves, and hurricane-driven disaster modification cycles.
The 90-Day Backlog
The loan modification process, on paper, is straightforward. A borrower falls behind, submits documentation, and a servicer evaluates eligibility across investor guidelines before issuing a decision and sending modification documents. In practice, the process routinely takes 90 days or longer.
The bottleneck is structural. Servicers typically support five or more investor types, including Fannie Mae, Freddie Mac, FHA, VA, USDA, and private labels. Each has its own guidelines and eligibility criteria. Underwriters often work through calculations manually, sometimes on spreadsheets, a process that can take 45 minutes per file and still produce errors. After decisioning comes document generation, shipping, borrower follow-up, executed document retrieval, and recording. Each step is often handled by a different team or vendor.
“That’s why it takes 90 days to get a loan mod through when the homeowner has submitted everything,” Bafna explains. “Somebody does the underwriting, then there are checkers checking the work, and then checkers checking the checkers.”
The human cost of that delay drives his work. “I always think about the honest homeowner who, due to death, divorce, or job loss, is going through this. How do we get certainty to those homeowners?”
Automation and Accountability
The push toward automation in mortgage servicing raises a question that compliance teams are quick to ask: where does automation end and human judgment begin? The answer, for default servicing, is not simply a matter of efficiency. It is a matter of legal accountability.
Underwriting decisions cannot be delegated to automated systems because every decision must be fully auditable, a standard that black-box AI cannot meet. Where automation does add value is in the supporting work: confirming calculations, flagging data inconsistencies, quality-checking returned documents, and moving completed files through to recording. “When you are in a compliance-driven world, you cannot say, ‘AI did this,’” Bafna says.
The broader implication is that compliance should not be treated as a late-stage review. Servicers and technology vendors that build regulatory requirements into the development process from the start are better positioned than those that treat compliance as a final check before launch. “Compliance stifles innovation only if you put it in two days before you launch something. If compliance is part of the journey, I don’t see them slowing anything down.”
The Data Barrier
Automation tools in mortgage servicing are only as effective as the data feeding them. Across the industry, that data remains fragmented and inconsistently formatted. Decades of legacy infrastructure built without standardized formats have created a foundational barrier that affects servicers, vendors, and borrowers alike.
The challenge surfaces most visibly when servicers attempt to connect new workflow tools to existing systems of record. Even when operational teams are eager to modernize, implementation stalls at the data layer. The problem is not a reluctance to change. It is the practical difficulty of extracting, normalizing, and transferring loan-level data across platforms that were never designed to communicate with each other. “The industry is talking AI, but it cannot even solve a simple API in some instances,” Bafna says. “Not having standardized data is the biggest challenge for the industry.”
Until the industry moves toward data standardization, the gap between what automation can deliver and what servicers can actually implement will remain wide. The conversation about AI-driven servicing cannot advance meaningfully without first resolving the more fundamental question of how data moves across systems.
Preparing for Defaults
Bafna believes the industry is approaching a period where default volumes will rise meaningfully. During the recent low-default period, rising home values allowed struggling borrowers to sell or modify without flooding the market. That cushion is now thinning. “We are going to get into that cycle,” he says. “When that cycle starts, if the default side is not using highly automated solutions, the industry will struggle to manage it.”
He also flags regulatory risk as a variable that servicers may be underestimating. CFPB enforcement activity has been limited under the current administration. Bafna cautions against assuming that will continue, noting that new leadership typically reviews past practices. Servicers operating on compliant, automated platforms will be better positioned regardless of which direction enforcement moves.
Reaching Borrowers Earlier
The default servicing process has a gap that neither technology nor policy has fully addressed: many homeowners who qualify for loss mitigation assistance never seek it. The problem is not a lack of available options. It is a lack of awareness, compounded by fear of consequences that are largely unfounded. Bafna estimates roughly half of eligible homeowners fall into this group, though he notes the figure is difficult to independently confirm. “They worry: if I call my servicer now, will they know I’m going to be behind? Will it affect my credit? None of that is true, but that is the worry.”
Bridging that gap requires reaching borrowers before they disengage entirely, ideally through channels that feel safe and low-stakes. Anonymous, self-serve tools that allow homeowners to explore their options without disclosing identifying information represent one approach. The goal is not to replace the servicer relationship but to prepare borrowers for it, so that when they do make contact, the conversation is more informed and more likely to result in a resolution on the first call. Earlier, better-quality borrower engagement improves the likelihood of a resolution on the first call, which benefits both the homeowner seeking certainty and the broader system designed to provide it.
For servicers, the window to address these gaps is open now. The conditions that have kept default volumes manageable are shifting, and the cost of being underprepared in a regulated industry is not only operational. It is reputational and legal as well.
About the Expert: Sapan Bafna is Founder and CEO of Outamation, a workflow automation company focused on the default servicing side of the mortgage industry, with nearly two decades of experience in mortgage servicing including a long tenure at CoreLogic (now Cotality).
This article is based on information provided by the expert source cited above. It is intended for general informational purposes only and does not constitute legal, financial, or real estate advice. Readers should conduct their own research and consult qualified professionals before making any real estate or financial decisions.
