Company: Canada Revenue Agency
Company background: Canada's federal tax administration and enforcement agency, with approximately 42,000 employees, the vast majority of which reside in Ottawa, Ontario.
Contact: Jason Andrew Oliver
Title: Senior Compliance Analyst and Data Scientst
Country: CANADA
Award Category: Innovative Problem Solver
Tell us about the business problem you were trying to solve?
In mid-2022, I was trying to solve the need for matching of TPD (third-party data) with individual SIN#s for T1 taxpayers - T1 being the CRA category for individual taxpayers, sole proprietorships, and partnerships (not corporate, which is T2). Manual matching was beyond tedious.
How did you use SAS to solve that business problem? What products did you use and how did you use them?
I used SAS DataFlux Studio v2.7, to create a workflow I called "AM/ER" (Auto-Matching / Entity Resolution). I set up an ODBC connection to our data mart, derived matchcode fields, and did the same with the TPD tables so I could join and fuzzy-match the two sides.
What were the results or outcomes?
I realized auto-match rates exceeding 50% in most cases, as long as the TPD had sufficient Address and Postal Code fields. Auto-matches could be done in mere minutes for a typical dataset. I evangelized the custom AM/ER method throughout my organization with an adoption rate of six org units.
Why is this approach innovative?
Since it hadn't yet been done at the CRA, and to that date I'd only done AM/ER with T2-Corp records (since 2019). I later added the DOB where a certain TPD table (e-funds transfers obtained by legal means) had DOB and so did our Data Mart.
