UCR Entity Resolution

Description

The UCR Programme's remit is to match and link records for Individuals Personal Tax, for sole traders and organisations VAT, Self-Assessment and Employer Payrolls. (4 of HMRC's key strategic heads of duty systems) to form a unique customer record. The Match Project, developed and deployed an in-house Enterprise Matching Tool (EMS) in November 2023. Currently, this capability allows for matching of individuals (without a business interest i.e., excluding self-employed individuals) using designatory customer data along with the key internal identifiers associated with the individual. EMS Entity resolution does not currently support the complexities associated with matching organisational data. Whilst, the in-house capability is able to link less complex customers/records relating to individuals using First Name, Last Name, Business Name, Tax Identifier, Address, and Date of Birth. The requirement for UCR's goes beyond this, not all heads of duty have an ID that can be linked across systems as is the case for organisational matching, sole proprietorships, partnerships. To develop the in-house capability beyond individuals would require significant work to address the limited functional features and encryption enhancements needed. That coupled with the significant cost to both run and maintain EMS no longer represents a return on investment. These finding have recently been presented to Technical Design Authority and a decision endorsed to replace the Enterprise Matching Tool (EMS) with a SaaS Entity Resolution Tool. Without an Entity Resolution capability, we cannot reduce the incidents of low confident matches, the associated risks and the impact of exceptions to operational areas. A SaaS Entity Resolution Tool will address these complexities and leave behind a future proof strategic capability through the UCR Programme that can deliver a performant, saleable and extendable matching service that can provide a view of complex corporate structures and legal entities e.g. Legal entity for a corporate's capital gains taxes might be different to the same corporations but different legal entities VAT. This activity can only be completed using an Entity resolution capability that goes beyond the current capability within Enterprise Matching Service (EMS), the solution developed in house. Effective Entity resolution will also support the business in addressing current challenges faced by the Operational Areas relating to Exceptions. Whilst a human can likely look at records from different sources and determined if they refer to the same product or not, this is not scalable across all records and the complexity increases quickly with more and larger datasets increasing resourcing pressure on operational areas. Effective Entity Resolution will work by reducing the instances of inconsistent matches, false negatives and false positives and derive canonical entities in a systematic and scalable way.