« Wisconsin Title Industry Veteran Going Strong at the Age of 90 | Main | Report: Business Ransomware Skyrockets »

07/10/2019

iTitle Agent: Automation and Artificial Intelligence Can Reduce Costs, Improve Efficiency

By Jeremy Yohe

Discussions around artificial intelligence (AI) and automation often range from apocalypse to utopia. The truth lies somewhere in between.
According to the Metropolitan Policy Program at Brookings, the power and prospect of these technologies initially alarmed experts, fearing that machine advancements would destroy jobs. The conversation now seems to be more balanced, accepting that robots are coming while also acknowledging there’s no need for scary dystopian alarms. The Brookings report found that about a quarter of U.S. jobs (36 million in 2016) face high exposure to automation in the coming decade. The most vulnerable include office administration, production, transportation and food preparation.

For title professionals, the emergence of artificial intelligence, automation and other technologies may seem daunting and alarming, but the solutions offer opportunities to help title and settlement companies meet evolving customer needs.

To understand how these technologies could help your operation, you need to know what each can accomplish. First, automation completes a specific task using technology—generally taking over repetitive and monotonous duties. There are plenty of these tasks in our world. Once called macros, they’re now known as robots. Meanwhile, artificial intelligence is the effort to mimic human thinking and reasoning using programs that write themselves via machine learning. To explain further, machine learning uses programs that update algorithms without additional human intervention based upon the analysis of large data sets to identify patterns of behavior, actions and results.

“The title industry is constantly automating more tasks. This isn’t new,” said David Floyd, chief data officer at NextAce. “What’s new is the application of AI to automation, which allows a huge leap forward in the things you can automate and how quickly you can automate.”

Marvin Stone, senior vice president of business integration for Stewart Title Guaranty, said this “Fourth Industry Revolution” is being driven by the combination of advances in computer technology and big data. He said cloud-based platforms opens the opportunity for more to utilize this technology.

“It would require a huge investment if we all had to develop these on our own, but they are available through various cloud providers,” Stone said. “As the title insurance industry more effectively collects, standardizes, digitizes and compiles data, it will allow for the enhanced application of AI and dramatic opportunities for automation.”

The application of artificial intelligence with machine learning is the next stage. Other lines of insurance are looking at AI in three big ways, according to Stone. These include setting pricing at individual levels based on risk, underwriting and analyzing data, and claims handling.

Stone said the claims process looks different in the title industry compared to other lines, but “we deal with this every day in title insurance, so we can expect to see application of this in a few years.”

Automation chart

In one example, video card manufacturer Nvidia is applying its facial recognition technology to collect data on homes following storms to find out the level of damage based off drone photography. Stone said this could be extended into the real estate and title space. Meanwhile, real estate agents are using autonomous cars to ferry people to home showings. Amazon’s suggestion engines have been around for years. Additionally, the suggestion engines employed by Amazon when shopping online are now being applied in the home shopping space.

“Automation is not about Terminator or HAL 9000—it’s really about bringing efficiency and addressing what you can do today and utilize technologies that are out there,” Floyd said.

Why Automate?

The chief reasons to automate tasks is to reduce costs and improve efficiency. Citing a Forbes article from 2018, Floyd said it costs 10 times more to run a title agency than it did 15 years ago.

“There’s huge pressure to reduce operating costs and figure out how to do things more efficiently,” he added.

Improving speed and accuracy of the transaction are other drivers for automation. “Every time a new technology makes turn times more effective, there’s greater pressure, and it becomes an expectation,” Floyd said.

An aging workforce and knowledge loss also fuel the need to automate certain tasks. Floyd said automation and AI can help memorialize the knowledge held by those retiring and spur the development of new ways to complete processes.

“As we look at automation and AI, we always must look at it through the lens of improving the customer experience through efficiency, quality and ease of use,” Floyd said. “Whether it’s the consumer, lender or real estate agent, we always want to make it a more efficient process. The use of technology is now an expectation and a driver for why someone will want to do business with you.”

Automation map

Where Is Industry Applying Automation?

The title industry has used automation in various parts of the process for many years. Here’s a look at a few:'

  1. Task and Process Tracking via Production Software (order through closing): Over the years, title production software companies have made great strides in automation that eliminates rekeying and the transferring of data from one application to another. This streamlines the order process and eliminates duplicative tasks.
  2. Automated Reconciliation: This allows title companies to reconcile accounts in a more timely manner and address issues more effectively.
  3. Workflow Management: As the industry improves automating tasks, it can start automating workflow a little more.
  4. Communications (automated emails, reminders, confirmations, notifications): Automating communications such as email are things that can be simple to implement in a title production system and are often overlooked.
  5. Title Production Automation: Floyd said we are starting to see quite a few players offering title production automation, including underwriters, third-party vendors and business process outsourcers.
  6. Commitment Preparation: The application of Optical Character Recognition (OCR), which is technology that enables different types of documents—such as scanned paper documents, PDF files or images—to be converted into other documents, continues to improve as the digitization of records has increased.
  7. Search and Examination: Searching is a pretty automated task that typically involves going into a database or title plant and pulling search results onto a screen or printout. The examination process follows.

Floyd said this is based on decision trees and a “very definable process with lots of ifs and buts. But through technology, you can automate more of the search and title exam process.”

“Some items such as communications are much easier to implement, but this goes to the fact that this is a continuous process in innovation and development to ensure we are all getting a little better,” Stone said.

Implementing Automation

Two terms people should become familiar with are bots (short for robots) and Robotic Process Automation (RPA). Nearly every repeatable computer task is open to using this RPA technology, according to Floyd. As an example, say a title company has a searcher who on a daily basis logs into a county website and enters a search that could include a name or legal description. After hitting Enter, results can be extracted.

“That whole process can be automated,” Floyd said. “Using RPA or other automation tools can deliver results back to another application.”

Many business process outsourcing (BPO) vendors already have applied these robotic processes to make searches more efficient and deliver faster turn times. Now, these companies are applying AI on top of the automation and making a leap forward on how quickly tasks can be updated.
Challenges with some automation tools include the ongoing management costs. Floyd said companies need to look at the cost of monitoring and programming for changes in procedures or source data versus labor to perform the task.

“It might be difficult to automate points that change constantly,” he said. Assessor data is one example. While there are automation tools that can capture this data, Floyd said assessors are notorious for updating information, and changing screen layouts and how data is formatted.

“This needs to be accounted for because now you’re switching some of your labor costs to management and programming,” Floyd said.

An item to think about is that frequent changes to source data and procedures will result in higher management cost and might drive implementation of technologies. The benefits include quicker turnaround, reduced labor management and more consistent procedures and quality.

“When implementing this in the title space, one of the things to think about in general is that most of the work we do is a pretty straightforward lot and block, subdivision type search on a resale or a refi where you aren’t going back that far,” Floyd said. “When looking at how to apply automation, start with the easy stuff. You can use RPA and pull in search results, and based on underwriting standards, you can deduct how many deeds back you need to go. Is the chain present? Are there additional names that need to be run?”

Stone said many title companies get hung up on the 100-percent rule versus the 80-20 rule. To help make automation more palatable, he suggested focusing on a particular segment of the business or a particular process that could be improved through automation.

“People are often unwilling to get started because they can’t do the big stuff,” Stone said. “Pick something and get a feel for it.”

Where Does AI Fit?

AI is already being used in many areas in the transaction, including:

  • Valuation models
  • Risk assessments
  • Workflow management
  • Fraud detection
  • Customer experience

Floyd said AI can identify the type of transaction based on search results. Running against algorithms, the complexity of an order can be determined and assigned to a specific person who’s best prepared to handle the transaction.

“One of the key areas helping to limit risk and workflow management is being able to direct a transaction to the best resource within your organization. In terms of automation, it allows a company to reassign an entire workforce from the mundane task-based processes and get them focused on revenue and risk. One of those is fraud. AI will play a major role moving forward in fraud prevention.”

Stone added that, through available technology and various vendors, title companies can bucket orders.

“The integration of more software with title production systems allows companies to analyze risk on a particular property or search and help them figure out how to manage the workflow and put it into the appropriate bucket, so the right things are being done by the right people.”

Despite all the benefits of automation and AI on the front end of the transaction with the search, Floyd says the industry must stay mindful not to forget about the curative process.

“That’s a large part of our lives and a reason why our datasets are so strong,” Floyd said. “We cure as we go along. As we look at title automation, that side of it can’t be missed and will be part of that risk profile. What curative actions need to happen and how do you automate the curative actions.”
Ultimately, the technology should lead to better customer service by analyzing the transaction on the front end. Interaction with customers can be tailored similar to the experience users have grown accustomed to Amazon—when you log in, it knows your trends and makes suggestions.
“That will cross into our realm in the next few years,” Floyd added.

The basis of artificial intelligence is really data—the accumulation of smaller data sets and bringing them together to paint a bigger picture. Floyd said that most documents are digitized at least back to the 1980s—and is constantly being pushed back further. Streamlining starter files could be one area to improve the process and be used to automate the search. Newer areas of application deal with consumer data, such as the types of deals they typically handle. Data sources include:

  • Digital documents
  • Public records
  • Assessor
  • E-recording
  • Starter files
  • Customer demographic data
  • Consumer profiles
  • Credit histories
  • Economic data
  • Transaction histories
  • Customer interaction histories

What’s a Title Agent to Do?

From a title agent perspective, Stone said this can become a big, daunting list. He said agents should reach out to vendors who can help and untangle different data sources. Floyd recommends examining each computer task to understand what you’re doing and how to prioritize. He said to start with easier transactions such as refinances and expand to other products.

Title agents also don’t have to develop these technologies on their own. Floyd suggested finding partners. The best advice is to complete a full review of workflow and processes, and then decide where improvements can be made.

“Everyone needs to pick up the phone and find out what their vendors can do. And tell them what you want to do,” Stone said. “Vendors are looking at this technology and thinking of ways to make title agents’ lives better. They are listening and want to provide automation where they can.”

Roadmap to Automation

  • Inventory Tasks: ­Review tasks you do every day
  • Prioritize: Don’t do everything at once
  • Educate: Develop internal subject matter expertise
  • Partner: Put their expertise to work for you
  • Experiment: Proof of concept/Pilot with customers
  • Execute: Plan then execute

Jeremy Yohe is ALTA’s vice president of communications. He can be reached at jyohe@alta.org.

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Working...
Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.

Working...

Post a comment