Dept. of Treasury (IRS) (IRS)

GTC consultants supported a project with the Dept. of Treasury (IRS) to help with their Automated Underreporter Program (AUR). AUR is a critical component of the IRS/Treasury strategy for reducing the individual underreported income tax component of the Tax Gap.

In support of the AUR team, we incrementally developed techniques, queries, and analytical models that highlighted historical patterns to select the most productive cases. The AUR application functionality we supported provided a more robust, less manual method for finding populations of cases prone to assessments, screenouts or no changes. With over 100 data fields with which to select cases, the matching of those data fields with the results for cases from prior tax years provided a deeper sample of millions of cases for building rules. Our work resulted in intensive sample gathering and new rule generation. The support we provided not only allowed new rules to be built but the effectiveness of existing rules to be tested. The new availability of data as a result of enhancements we made to the system allowed the AUR team to create highly detailed rules for the selection of cases. For instance in prior years, without doing any sort of in depth sampling, performance could only be monitored at the microcategory level, with some microcategories potentially including hundreds of thousands of cases.

With the enhancements we implemented to the AUR application, rules can be created at a much more detailed level. The business rules are based on the most recent business intelligence and are meant to automatically select productive cases and automatically non-select less productive ones, so those cases bypass the possibility of selection based on their Estimated Potential Assessment. The incremental addition of business rules twice per year served to hone the selection of cases and thereby provided the most productive inventory to the program.

Back To Projects

Recent Projects