MBA NewsLink Q&A: Verity Global Solutions’ CEO Sam Mehta on AI Fatigue

Sam Mehta is the founder and CEO of Verity Global Solutions, an outsourcing partner that helps lenders scale production capacity without expanding fixed headcount. With teams based in the U.S. and India, Verity combines trained mortgage specialists, embedded automation and disciplined process design to support functions across the mortgage lifecycle, including origination operations, title production, accounting, QC, and servicing support. Before entering the mortgage industry, Sam worked in advanced manufacturing, where he developed a process engineering approach to operational excellence.

MBA NewsLink: Why do so many AI and automation initiatives fall short, despite significant investment?

Sam Mehta: Most initiatives fall short because they are applied to processes that are messy or unstable to begin with. For example, if a lender has three different teams collecting borrower data three different ways, automation will simply speed up the confusion and create more downstream risk.

Many lenders also have unrealistic expectations about ROI. Anytime you implement new technology, it should be tied to a specific outcome. Yet many organizations invest in AI or automation initiatives without a clear operational goal in mind. When the initiative delivers mixed results or creates more problems, whatever the lender spent in time and money becomes a sunk cost.

Yet another issue is the tendency in our industry to chase the latest shiny object when it comes to technology, which can be very costly. There are many exciting new technologies, but because they are new, they are often unstable and carry unexpected maintenance and training costs. Most lenders would be better off waiting for certain tools to mature—or at least outsourcing to experts who have already done the hard work of applying them to specific processes in ways that create real value.

MBA NewsLink: What is process redesign, and why is it important when implementing new automation?

Sam Mehta: Before adding technology to a specific process, the process itself has to make sense. It’s like organizing a workspace before you install new equipment. If you have files scattered everywhere, whatever equipment you bring in is less likely to improve your productivity and more likely to create a bigger mess.

Process redesign involves standardizing how work is done, setting checkpoints, and removing extra steps, so that automation drives scale instead of creating more work. Many times, we have seen lenders try to automate onboarding or QC workflows before mapping them out, only to end up with more exceptions and compliance issues. But when you take time to redesign the process first, automation scales the right things, like accuracy, speed, and cost control. Without that foundation, automation often becomes just another layer of complexity.

MBA NewsLink: How can lenders evaluate which tasks are truly automation-ready versus those that still require human judgment?

Sam Mehta: Automation is ideal for tasks that are rules-based, repeatable, and driven by structured inputs. If the work follows a clear pattern—such as pulling data from standard borrower documents or checking for missing signatures—there are many tools that handle these tasks easily.

But again, you have to work on the process first, because the more processes you can standardize, the more they can be automated.

Activities that are not typically “automation-ready” include clearing loan conditions, reviewing appraisals, underwriting non-traditional loan products, and working with borrowers. These tasks still require judgment and oversight by human experts because they involve context, interpretation, and risk awareness, which are areas where automation falls short.

Generally speaking, lenders get the best results when they leverage automation to handle repeatable processes and let people focus on problem solving and borrower communication.

MBA NewsLink: What role do trained experts play in making automation sustainable over time, especially as volumes fluctuate?

Sam Mehta: Sustainable automation initiatives depend on people who know the work and understand risk, and can intervene when needed. Automation, while powerful, is unable of thinking for itself or adjusting to market conditions. When volumes rise or fall, it takes trained experts to keep operations stable by managing exceptions, fine-tuning rules, and helping lenders absorb new volume without breaking SLAs or controls.

When the next refinance surge hits, for example, automation will help teams index files and move documents faster. But when guidelines or loan programs change, someone still needs to adjust a lender’s rules and make sure nothing slips through the cracks. During slower periods, you need those same experts to focus on improving workflows and preparing systems for the next cycle.

MBA NewsLink: What practical steps can operations leaders take to reduce “AI fatigue” and refocus teams on measurable efficiency gains?

Sam Mehta: AI fatigue is real, and it’s being fueled by frustration with adopting AI tools through trial and error as well as by industry hype. Today’s lenders are being flooded with bold claims about AI from every direction. This places pressure on them to act quickly while also creating anxiety about making the wrong choice. In many cases, lenders don’t understand what certain AI tools do or how they work. When lenders introduce a new tool and their teams don’t see any workflow improvement, frustration builds quickly.

To prevent this from happening, leaders should keep things simple. Start small, and use AI or automation on repeatable tasks where you can measure the results. At Verity, we measure success in saved labor, reduced errors, fewer touches per file, and closing speed. It’s equally important to celebrate quick wins and clearly link each effort to something teams can see and quantify.

In many cases, especially for small and mid-sized lenders, the smart approach is outsourcing with a partner that has already built up expertise at applying AI and automation to specific mortgage processes in ways that actually ease workloads. When lenders and their teams see AI and automation taking real work off their plate, their confidence improves along with their bottom line.

(Views expressed in this article do not necessarily reflect policies of the Mortgage Bankers Association, nor do they connote an MBA endorsement of a specific company, product or service. MBA NewsLink welcomes submissions from member firms. Inquiries can be sent to Editor Michael Tucker or Editorial Manager Anneliese Mahoney.)