AI in Title: The Future Is Now

AI in Title: The Future Is Now

Automation has come a long way in the settlement process. But new AI approaches may finally tackle some of the oldest and most stubborn chokepoints plaguing agents, underwriters, Realtors and consumers.

ISSUE 100 | VOLUME 6 | June 2021

AI in Title: The Future is Now

THE TITLE INDUSTRY IS RIPE (AND HAS LONG BEEN) FOR NEW SOLUTIONS to age-old challenges. It’s no secret that the settlement process, regulated on a state-bystate basis and subject to varying lender or underwriter requirements, is not unlike the original Frankenstein’s monster: bolted together pieces and parts. It’s a process replete with numerous manual tasks or redundant actions. In fact, you’d be hard pressed to find someone within the industry to say the settlement process, as it stands, is a shining example of efficiency.

Agents and underwriters alike have long sought ways to streamline their workflows, and although we have made encouraging process in recent years, there’s more to be done. Much of the progress made has come in the form of automation. So the tasks or processes most easily adaptable to simple automation—usually in areas involving simple, one or two step processes, like document preparation, data input, data retrieval and the like—have gotten the most attention first.

We’ve already seen great automated solutions helping agents with several of the simple functions in production; elements of closing such as remote online notarization technology and some limited communication functions.

One type of automation tech we’re starting to see more often is Robotic Process Automation (RPA), which is sometimes confused with AI. RPA can be used to repeat multi-step processes. However, unlike AI, RPA’s performance depends entirely on how it’s programmed. While AI can “learn” from some mistakes, RPA will repeat mistakes (garbage in; garbage out). Nonetheless, it’s extremely useful in automating more complex actions.

And yet, the traditional settlement process, from title through post-closing, also requires numerous, multiple-outcome processes that aren’t as easy to automate. Order intake or data retrieval from consumers come to mind. It’s the elements involving human beings that can be the most resistant to streamlining. For example, consider the chokepoints in the typical title agency workflow that require communication or collaboration among parties (buyer, seller, real estate agent, closer, loan officer), with all of the complexities and complications that arise. We still have yet to see as much progress in streamlining there. Look at, for instance, the wide variety of ways a simple sales contract comes into a title agency, from handwritten to PDF; coming in via fax, courier or even by hand. All of that needs to be processed into the same system, and it’s the title agency’s job to decipher those agreements.

While there have been, for example, multiple efforts to forge a “communication portal” where all parties can come together to exchange information or documents, too many have ended up only creating another silo and, yet another, stand-alone solution that doesn’t integrate itself into the global workflow well enough to stop being a chokepoint.

The age-old approach to speeding things up for a more efficient close, so far, has been to “staff up” with more humans to handle those tasks. Simultaneously, tasks requiring collaboration and good communication, even if redundant or simple, have also tended to be managed by humans using ill-fitting tools like email or telephone. It’s here where we see, even today, some of the greatest opportunities for error, delay or even fraud. The common but obvious complaint about data or information getting lost in a flurry of emails or voice mails is very valid. And yet, many agents still rely on their team members to do “whatever it takes” (emails, texts, calls) to share the status of the transaction with the buyer’s real estate agent. It’s humans, rather than tech, that are usually charged with figuring out how to prepare closing documents like the CD from piecemeal closing instructions; or determine tax proration from county to county. We’ve heard that a title agent’s job is to stand at the bottom of a mountain; watch a snowball become an avalanche on the way down and then catch the contents and plug them into the closing process. Every real estate agent deals with a title agent differently. Every agent transmits documents and data differently. So, finding a technology that allows agents to catch every bit of that avalanche can be a challenge.

Until now, most owners and title agents have had little recourse but to stare down the inefficiencies, shrug their shoulders and say, “It’s the way we’ve always done it.”

However, the emergence of new AI technology developed just for the complicated title process may be one of the most promising developments we’ve seen in the ongoing quest for modernization in quite some time. AI naturally lends itself to multi-step processes that require fairly complex decisions from point to point. Already, we’ve seen it put to use effectively for processes requiring the parsing and extraction of select data, then routing it to one of a number of options. It’s also becoming an important means of taking manual tasks that require some judgment (e.g. “stare and compare”) and making them more efficient.

But perhaps the greatest potential use for AI will come in the simplest—yet most difficult to attack—challenges: interparty communication. Just think of how many times and how many ways a traditional title staff needs to hop on the phone or send an email (or three) to, say, get a copy of a driver’s license; or gather information from a loan officer or real estate agent; or explain to the borrower, two or three times, that a closing will really happen on the scheduled day. We’re already seeing new AI-based technology that allows title agents to communicate with interested parties and clients without requiring two missed calls, three voice mails and an email to exchange one simple bit of information.

It’s the flexibility of AI and its ability to make multi-faceted decisions that may allow us to finally use fewer human resources to handle otherwise simple and redundant tasks made complex solely because they involve human-to-human interaction. Now, it’s also time for title agents and owners to take a hard look at their own workflow processes. Any technological upgrade starts—and ends—with strategy and leadership. Examine your production process as objectively as you can. Where are the biggest wastes of time and money? Never mind that there may not yet be a precise, comprehensive tool to solve it end-to-end yet. Are there solutions that tackle part of those tasks? Can you negotiate customized integrations or other modifications from your technology partners that better integrate your workflow? It’s one thing to employ technology to save time and cost. But too often, the result of a poorly planned or ill-fitting tech deployment is front-line workers working from two monitors, with five open browsers or programs, glancing at a couple of apps on their phones or tablets continuously—until they need to get back on the phone or send an email.

Technology developers are always looking to provide better answers to decades-old questions. With AI, we can see new possibilities. But the people developing new solutions also need to hear from the real experts on workflow and chokepoints—title agents and professionals—to perhaps finally take us another step closer to a true end-to-end settlement technology. Let them know where the process needs to be improved. Finally, the future is here for the title industry. Let’s meet it half-way.

Read more: The Time for Title Standards is Now