Zementis: Going beyond the rules

In a previous post, I pointed out that BRM (business rules management) is going to be big in 2006. Lenders will find that managing their guidelines in an automated fashion will pay huge dividends. But at least one firm is hoping that some lenders are willing to go one step beyond that.

Zementis, Inc., San Diego, is marketing the Jade automated underwriting solution. Not an AU system, per se, though it appears that it could serve as a standalone solution, the tool is designed to sit above the lender's AU system, catching exceptions and performing analytics that company CEO Michael Zeller says will let the lender get more out of its automation.
"The complexity of nonprime/high-risk deals is difficult to automate with a rules system and the maintenance of the rules is not trivial. The addition of predictive analytics provides a significant boost to the effectiveness of the AU system. Furthermore, the risk assessment of each loan provides benefits to other aspects of the business process, e.g., quality control, smart queuing, and pricing," Zeller said in an e-mail.

Over the phone, he explained that Jade's analytics probe back into the lender's own portfolio, digging up information on how other similar loans have performed in the past and how profitable they were for the lender. It then uses this data to try to reach an automated decision when the legacy AU throws an exception.

Zeller says his new company (they've been in business less than 2 years) has developed a Model Development Life Cycle that allows it to develop, test, and deploy predictive models in an iterative way.
"The client can gradually 'tune' in the JADE system, which is important since the system has direct impact on revenue and profitability. Hence, we can start very conservative and extend its automation step-by-step, over time. "

Obviously, the more data you have the better the analytics will be at reaching valuable conclusions, so Zementis is targeting larger lenders now. Accredited Home Lenders Holding Co., also based in San Diego, has already agreed to deploy the system, according to a company press release.


Popular Posts