Skip to main content

Coester Announces New AI Appraisal Review System

NationalMortgageProfessional.com
Oct 28, 2013

Coester Valuation Management Services has announced the release of its new automated artificial intelligence based appraisal review system. Appraiser Intelligent Review, or AIR, employs appraiser based logic to automatically review all reports completed by CoesterVMS network appraisers. The two-part process includes an initial validation of key fields on the report followed by nested analytics to ensure adherence to applicable FHA and GSE requirements. “Nothing like this has even been close to done before, we are not using just a base checklist but this is actual appraisal logic and will be more accurate than anything ever done," said Brian Coester, CEO of CoesterVMS. "The great part about this is the more information and data a client puts into the system the more accurate it is. In certain geographic areas where we have a lot of relevant data the report is as accurate as an appraisal desk review because we have data from hundreds of reports in the local market area to compare the appraisal report to." CoesterVMS developed basic Appraisal Validations to eliminate time consuming conditions relating to transposed verbiage or minor errors. Immediately upon report transmission, the Cloud Control System provides appraisers notification regarding any fields discrepancies between order data and the uploaded report. Following validations, the system scans an XML version of the appraisal against two million lines of code and over 150,000 algorithms to tract and score appraisals based on actual appraiser logic.  XML review results in a scrub against rules representing the tenants of regulatory and investor valuation guidelines.  In the event a report exceeds allowable logic for a specific rule, additional logic factors in specific market conditions and market adjustments to identify support for the discrepancy. Clients receive a detailed breakdown of all logic employed by the system, as well as commentary from quality control staff in the event the system recommends further analysis.
Published
Oct 28, 2013