I’m working with a large mortgage note portfolio and trying to create a more reliable way to identify data problems before the assets move into deeper due diligence.
The portfolio contains fields such as loan balance, payment history, interest rate, maturity date, property information, lien position, tax status, and assignment records. The challenge is that some records are incomplete, some values do not match across different files, and a few assets appear to have unusual payment or lien data.
I’m interested in using data analysis to find these exceptions automatically rather than reviewing every record manually. For example, it would be useful to identify missing values, duplicate records, unusual loan balances, inconsistent dates, or assets where the reported lien position does not match another data source.
For those who work with JMP or similar analytics tools, what approach would you use to build a repeatable workflow for this type of mortgage note data-quality problem?
Would you use statistical outlier detection, data validation rules, visual dashboards, or another method to identify the records that need deeper review?
