Cenlar is the largest mortgage and servicing company in the US. The company essentially offers mortgage servicing as a service for clients. Some clients outsource servicing to Cenlar, and other customers use the company's services behind the scenes under their own brands.

I caught up with Josh Reicher, Chief Digital Officer of Cenlar FSB, to talk about the business and technology.

The business. Mortgage servicing is known for thin margins and enterprise IT, automation and efficiency are critical. Reicher said Cenlar's "focus is on execution so we can be best in class." "If we are minimizing mistakes and doing the best we can in terms of automation we're able to deliver better services and more profitability to shareholders," he said.

The approach to automation. "When we look at automation the first thing, we look at right now is risk. How can we decrease our operational risk? Everything follows when we decrease risk. We will improve our performance. We will improve our profitability. We improve the experience for homeowners. And we help our employees provide better services," said Reicher.

Reicher said Cenlar said hyper automation and intelligent automation is big for Cenlar. The aim is to provide an automation toolkit that can solve different problems across the company. "If it's just a repetitive process that needs to be automated standard RPA can get the job down once you script it out," said Reicher. "If it's going to require more intuition, we can start to look at the applications of AI. We're looking at AI now for natural language understanding and starting to get into generative AI."

Generative AI. Reicher said Cenlar has a few proofs of concept in the works, but the company has to have the right controls and safety in place first. "We have to make sure that our clients and their homeowners are safe before we release this technology," he said. "First we're going to make sure the data never leaves our confines. It's an obvious one, but something we state. We're in a regulated industry."

"It's also very important that any type of technology or AI delivering dynamic intuitive responses never goes directly in front of a homeowner without a human in the middle," explained Reicher. "The idea is that a back office agent is looking at a response and decision and seeing that they are accurate. If accurate, it will be sent out. We're enabling the back office to provide better services than create risks."

Vendors for generative AI. Reicher said that Cenlar is talking to several partners because time to market is important. "Right now, it's really important to build up that generative AI competency internally," said Reichert, adding that his team has strong developers and architects to design internal solutions.  Ultimately, generative AI would be integrated with workflow automation.

Build vs. buy. Reicher's approach is to avoid software that's best-of-breed and focused on one specific function. "The problem of using a point solution is that eventually you have enough of those one specific function providers that you have too many," he said. "As an IT organization you have to support all of the technology and you have to make sure that it is enabling the business instead of slowing it down."

Return on investment. "First of all, the returns have to be quantitative. If returns cannot be measured it's not going to get continued investment," said Reicher. "The first return is can we decrease mistakes right away. There are usually penalties associated with mistakes. We also find efficiencies in elevating employees to more analytical and thoughtful roles, say problem solving and supervising vs repetitive things."

Controls are another key area for ROI. If you can automate controls then you improve risk scores, said Reicher. "When we're implementing automation, we look at the short and long-term results. That all adds up to profit and loss impact," said Reicher.

There's also ROI in using data to be more predictive. Since Cenlar services loans, it can use mortgage originations to prepare for future demand and scale up and down.

Looking ahead. Reicher said part of his role is solving business problems right now, but a lot of it is educating the executive team so they understand how the technology roadmap will set up. "Our board and our executive team want to go two to five years out from now so we can start preparing," said Reicher. "We know AI is going to be a very big deal for our industry and that's why we're building out that capability now. We have an idea lab we've been running for years now trying to stay ahead of the curve."

Related: